We are usually about forty minutes in.

By then we have been through the photograph, argued gently about the headline and rewritten the first two lines of the About section three times. The person on the other end of the call is engaged, taking notes, pleased with the progress. This is the enjoyable part of my week :)

Then I scroll back up to the top of the profile and ask what looks like an administrative question. What have you got in your Industry field?

There is a pause. I can hear the click. And then, almost every time, some version of the same sentence.

“Oh. That says Law Practice. I left law in 2019.”

Or teaching. Or the agency they sold. Or, more often than you would think, an industry they never worked in at all, because they were setting up the profile in a hurry and picked whatever the dropdown offered first.

What follows is my favourite part of the conversation, because it is the moment somebody realises they have been doing everything right and one small thing wrong, and the small thing has been costing them more than any of the big things.

When a buyer builds a prospect list by sector and you are filed under the wrong one, you were never a candidate for that list. Nobody writes to tell you. Your profile still looks perfectly good to you, because you are the one person who never sees it through somebody else’s filter.

You can spend three years posting consistently, refining your headline and growing your network, and stay invisible to an entire category of people who were searching for exactly what you do. The dropdown took four seconds to fill in, and it has been making that decision on your behalf ever since.

This post is for those who already have a LinkedIn profile set up but are not getting as many inbound approaches as you would like.

Here is what to do about it.

Subscribe now

What actually changed

LinkedIn used to run a fairly short industry list. A few dozen broad buckets, written in general language, most of which described a whole sector rather than the work anyone actually did.

That list was replaced. The current version, known as Industry Codes V2, is mapped to the North American Industry Classification System (NAICS), which is the standard statistical agencies use to categorise businesses.

Instead of a flat list of broad options, it is now a hierarchy: 434 industries sitting under 20 top-level sectors, branching down through several levels, each with its own code and description. The reference table is published by Microsoft, which owns LinkedIn (Microsoft Learn, Industry Codes V2).

The list you chose from is not the list that exists now. If you set your industry more than three or four years ago, you picked from a menu that no longer exists, and the option that describes your work may have been created after you stopped looking.

The example that makes it concrete

Take two entries sitting side by side in the current reference table, both under Financial Services, both under Capital Markets.

Read those twice. They sound like the same job to anyone outside the sector. They describe two different businesses, two different regulatory positions, and two different sets of buyers.

An adviser, a wealth planner or anyone whose work is client advice for a fee sits in the first. A discretionary portfolio manager, a fund manager or someone in investment operations sits in the second. Choose the wrong one and you appear in the wrong searches, run by the wrong people, looking for something you do not do.

Now multiply that across every sector.

Professional Training and Coaching sits separately from Education Management and from E-Learning Providers.

Software Development sits separately from Information Technology Services and from Technology, Information and Internet.

Marketing Services sits separately from Advertising Services and from Public Relations and Communications Services.

Each of those distinctions looks like hair-splitting in a dropdown menu. Each describes a different market.

I have built you a tool for this

The reason people pick the wrong option is that the LinkedIn dropdown is a poor way to explore 434 choices. You type two words, glance at what appears, and take the first thing that looks close.

I have built a free searchable version of the full list: The LinkedIn Industry Finder.

Check your Industry is Correct

Search a term or filter by sector, and it shows you every matching industry with its position in the hierarchy, so you can see what sits above and beside your option before you commit to it. Seeing Investment Advice and Investment Management next to each other, both under Financial Services then Capital Markets, tells you more in five seconds than the dropdown will tell you in five minutes.

It is a reference tool. It does not touch your profile, and you make the change yourself on LinkedIn, which takes about a minute. Trevisan Consulting is not affiliated with or endorsed by LinkedIn Corporation.

What the Industry field actually does

Four things, and it is worth separating what is documented from what is inferred, because plenty of LinkedIn advice does not bother.

1. It filters you in or out of searches. This one is not in dispute. Industry is a standard filter in LinkedIn’s own search, in LinkedIn Recruiter and in Sales Navigator. LinkedIn publishes guidance for recruiters on filtering candidates by industry (LinkedIn Recruiter Help). When somebody builds a list, whether they are hiring or prospecting, industry is one of the boxes they tick, and profiles outside the selected industry never enter the results.

2. It works hierarchically, which cuts both ways. Because the list nests, selecting a broad parent category can pull you into results for that whole branch, while selecting a precise sub-industry makes you findable by people who filter narrowly and invisible to people who do not think to expand. Neither is automatically better. It depends on whether the people you want are searching wide or searching precisely, and in most professional markets the valuable ones search precisely.

3. It frames you at a glance. A human reading your profile takes the industry as a statement of what world you belong to. When your headline says one thing and your industry says another, the reader has to decide which to believe, and busy readers do not usually decide in your favour.

4. It may affect distribution. Various third-party sources claim LinkedIn uses industry as one input when deciding who sees your content. That is plausible and widely repeated. I have not seen LinkedIn confirm it directly, so treat it as a reasonable working assumption rather than an established fact.

The first two alone are enough reason to spend ten minutes on this.

The four ways people get it wrong

In the profile reviews I run, nearly every industry problem falls into one of these four. The last one is the dangerous one, because it looks correct even when you go and check.

The second row is the only one that is a genuine judgement call rather than a mistake. If you want more work like your current work, choose the discipline. If you want to be found by the sector you serve, choose the sector. Choosing on autopilot is the error, not choosing either one.

The fourth row is the one nobody catches, because nothing looks wrong when you open your profile.

How to choose, in five questions

Work through these honestly and the answer usually resolves itself.

That last question is the one that unsticks people. You are not describing your entire career. You are describing where you want to be found now.

Here is how it plays out for the situations I see most often.

None of the right-hand options is automatically correct for you. Each one is a prompt to go and look, because the left-hand column is where the old list would have put you and the right-hand column is where the current list might.

If you run a business

There are two separate decisions here, and running them together is a common and expensive muddle.

Your personal profile industry describes you.

Your Company Page industry describes the business.

They can differ, and often should. A founder who came from finance and now runs a software company might reasonably sit in Software Development on both, or in Financial Services personally if that is the market they sell into and the world they want to be known in. There is no universal right answer, and there is definitely a wrong one, which is never having thought about it.

Three things worth checking today.

1. Your Company Page industry is what LinkedIn uses to place your business in searches and in its own competitor comparisons. If it was set by whoever built the page in 2018, it is probably still the broadest option they could find.

2. If you sell to a specific sector, your personal industry is one of the ways buyers in that sector filter when they build lists. Business development people use exactly the same filters recruiters do.

3. If your team members all have their industry pointing somewhere different from the company, your firm’s presence on LinkedIn looks less coherent than it is. That is worth ten minutes in a team meeting and a shared instruction.

If you are, or might one day be, looking for a job

Almost everybody falls into the second half of that sentence, which is the reason to fix this while you are comfortable rather than while you are urgent.

You are happy where you are. Set your industry to the work you actually do, at the most precise level that is honest. You are not signalling anything to your employer by doing so, and you make yourself findable for the approaches you would want to hear about rather than the ones you would not.

You want to move sector or discipline. Point the industry at the market you are targeting, provided the rest of your profile makes that credible. Somebody with a legal background moving into financial services compliance can reasonably select the financial services sector, as long as their headline, About section, experience and Featured content show relevant regulatory and risk work behind the claim. Industry is a supporting signal. It will not do the positioning on its own, and a profile where the industry is the only thing pointing at the new direction reads as wishful.

Make the rest of the profile agree with it

A correct industry sitting next to a stale headline achieves very little. Once you have chosen, check that four things point the same way.

Your headline names the work, in the words the market uses for it. Your About section opens with what you do rather than a chronology. Your skills support that same discipline, with the strongest ones ordered at the top, because skills are searchable too. Your experience descriptions use the vocabulary your industry implies, in outcomes rather than duties.

You may find this guide a sueful starting point

Access Guide

When all five agree, a recruiter, a buyer or a potential partner reaches a conclusion about you in a couple of seconds. When they disagree, that person has to work out which version of you is current, and most of them will not bother.

How to change it and it takes a minute

First, find your option.

Open The LinkedIn Industry Finder, search your discipline or your sector, and note the exact wording of the industry you want. Do this before you go anywhere near the profile editor, so you are choosing rather than settling.

Then, on your personal profile

1. Go to your profile.

2. Click the pencil icon in your Introduction section, the top card holding your name, headline and location.

3. Scroll to Industry.

4. Start typing the area you want, rather than scrolling the whole list. The search reveals options you will never find by browsing.

5. Read the options that appear before picking. This is the step everybody skips, and the entire point of this article.

6. Save.

   

On your Company Page, the industry sits in the page admin settings under the page information. Same principle applies. Type, read, choose precisely.

One caveat before you change it

Your choice has to survive contact with the rest of your profile. If somebody clicks through from a search and finds nothing that supports the category you have placed yourself in, you have bought a click and lost the credibility that was the point of getting it. The right industry is the most precise honest description of the work you do and the market you want to be found in. Nothing beyond that.

(A note on free versus paid LinkedIn search: Industry is no longer as accessible as a people-search filter to ordinary unpaid members as it once was. This means the direct discovery effect may be more limited for general LinkedIn users. However, Industry remains a search filter in LinkedIn Recruiter, Recruiter Lite, Talent Insights and Sales Navigator - tools used by recruiters, talent teams, sales professionals and business-development teams. It is therefore best viewed as one accurate supporting signal in your wider positioning, rather than a standalone optimisation tactic.)

The one thing to take away

You will spend hours agonising over a headline that maybe forty people read closely this year, and four seconds on a dropdown that decides whether you appear in searches at all.

Go and look at it now. Open The LinkedIn Industry Finder, search the work you actually do, and compare what comes up with what your profile currently says. If it is already right, you have lost a minute. If it is wrong, you have just made yourself findable to people who have been filtering you out for years without either of you knowing.

If your team’s profiles all point in different directions, that is usually a twenty-minute fix across the whole firm, and it is the sort of thing we cover early in LinkedIn training because it costs nothing and changes who sees you.

Over to you. Go and check your industry, then tell me in the comments: was it right, was it stale, or did you find an option you did not know existed?

Subscribe now

Connect with me

Substack | LinkedIn | YouTube | Website | Facebook | Instagram

3,237,176.

Yes, 3,237,176.

That’s how many impressions LinkedIn’s own analytics say I generated in the last 365 days, up 607% on the year before, from 348,673 impressions to 2,467,447 across my own public posts.

When I decided to write this last week, it was an idea. It turned into a magnet opus as I pulled every post I’d published with real performance data attached and started looking for a pattern.

The growth over the last six years wasn’t one lucky post. The monthly trend shows a steady, compounding climb through late 2025 into 2026, not a single spike:

And the year-over-year jump is stark when you put the two 12-month periods side by side:

What I found when I dug into why wasn’t the pattern I expected.

The obvious place to start was my best posts so I sorted by impressions and looked at the top of the list.

A post about a rude message in my inbox: 233,472 impressions. A birthday reflection: 206,330. A story about a pay-rise negotiation: 145,611. Christmas abroad: 94,608.

I found that just 34 posts, 3.4% of everything I’d published, accounted for 47% of all my impressions across the entire dataset. If I’d stopped my analysis here, the article would have written itself: “Here are my viral posts, go copy them!”

But that would have been the wrong lesson.

When I checked what those 34 posts actually did beyond reach, most did surprisingly little.

• The birthday post, my second-highest by impressions, brought zero recorded profile views.

• The holiday post reached 94,608 people and isn’t something I can repeat on demand. I didn’t plan for it to go anywhere near that far. It just did.

The boring middle that actually built the account

Here’s the number that changed how I read my own data: 590 posts, 60% of everything I published, sat in what I started calling the “boring middle.” Not viral, not ignored. Between 3,000 and 20,000 impressions each.

Individually, none of those posts is a headline. Added together, they account for 46.6% of my total impressions, almost exactly matching the 34 whales, from 17 times as many posts.

That’s the real story.

Unlike the whales, the boring middle is not random. When I broke it down by content format and opening-line hook, clear patterns emerged, patterns I could point to and say, do this again on purpose.

Format matters more than most people think

• Image posts were both my most-used format and among the strongest performers by average impressions and the only format that reliably drove measurable profile views (8.6 per post on average).

• Every viral post included a real photo of me

• Video and document/carousel posts also outperformed plain text and polls consistently.

The opening line changes everything

Posts opening with a specific number or statistic averaged 6,804 impressions, nearly three times posts that opened with a question (2,345).

Personal “I” statement openings were the second-strongest hook at 5,333 average impressions. Numbers and personal stakes consistently beat vague curiosity hooks and questions.

Where profile views actually come from (and it’s not where I expected)

I also wanted you to know which posts brought the right kind of attention, not just the most of it. I checked what correlated with profile views, the moment someone gets curious enough to click through and find out who you are.

Impressions and profile views were only loosely connected. A post could reach 200,000 people and generate zero profile views, or reach a fraction of that and quietly outperform. My highest profile-view post, the Christmas travel story, reached less than half the impressions of my top post, yet drove 2,289 profile views, by far the most of anything I published all year.

Across the full dataset, my 987 posts generated 4,014,179 impressions in total, which translated into 4,847 profile views, a conversion rate of roughly 0.12%. Small, but not nothing: those profile views, in turn, correlated far more strongly with new followers (0.84) than impressions ever did (0.63). Reach gets you seen. Profile views are what get you remembered..

Know who you’re actually writing for

Before applying any of the formats below, check who LinkedIn says is already engaging with your content.

My audience skews toward senior decision-makers, 23% senior-level, 11% director, 8% CXO, concentrated in financial services (16%), consulting (5%) and legal/professional services, with London accounting for 20% of the audience and smaller but meaningful clusters in Geneva, Zürich and Scotland (who knew?!).

YOUR TASK RIGHT NOW: Open LinkedIn’s own Analytics tab (not a third-party tool) and check the Audience Demographics panel before writing your next ten posts. If your content is aimed at one audience but your actual readers skew toward a different seniority or industry, your hooks, language and examples need to shift to match who is already there, not who you assume is there.

So what actually predicted the boring middle’s success?

Once I stopped chasing the whales and started studying the middle, five repeatable patterns emerged, not tricks, just things I could point to across dozens of posts and say “this, again.”

1. The number-led stakes hook

Why it works: Posts opening with a specific number or figure averaged 6,804 impressions, nearly three times posts that opened with a question. A number signals concreteness before the reader has invested any attention.

Template:

[Specific number, cost, or statistic] because of [a decision, mistake, or moment].

Here's what happened:
[3-5 short lines of narrative]

The lesson:
[One clear, transferable insight, not a generic platitude]

AI prompt:

Act as a LinkedIn ghostwriter. Take this story: [paste your story/anecdote].
Rewrite the opening line so it leads with a specific number, cost, or
statistic that creates curiosity. Keep the rest of the post under 150 words,
in short punchy lines, and end with one transferable lesson, not a hard CTA.

2. The personal-stakes story

Why it works: First-person openings (”I’m 50 today,” “I am not an activist”) averaged 5,333 impressions and produced my single highest-engagement post of the year (1,954 engagements), a post that sat in the whale bracket but still followed a format I could deliberately repeat.

Template:

[A personal moment or milestone, stated plainly]

For [a period of time], I assumed/believed [an old belief].

What actually happened:
[2-4 honest, short paragraphs]

[One line that reframes the lesson for the reader, not just for you]

AI prompt:

Write a short, honest LinkedIn story about [describe the personal moment].
Use first person, short sentences, no corporate language. End with a
reflection a professional audience, not just friends, would find
relatable. Do not add a sales pitch or CTA.

3. The single-client, single-lesson case study

Why it works: One post about a real client interaction (”The tale of Henry”) produced 82,940 impressions and strong, steady engagement, no viral hook required, which is exactly why this format belongs in the boring middle rather than the whale pile.

Template:

[A workplace or client situation, stated as a short scene]

What went wrong / what most people miss:
[1-2 sentences]

What I'd do differently / what actually worked:
[2-3 concrete, practical steps]

[Optional: a direct question inviting the reader's own experience]

AI prompt:

Take this workplace observation: [describe the situation]. Turn it into a
LinkedIn post that reads like a short story, but ends with a clear,
practical takeaway a business audience can apply. Keep it under 130 words.
Avoid generic advice, be specific to the situation described.

4. The single-image, high-caption-density post

Why it works: Image posts were both my most-used format and my strongest performer by average impressions (5,032), and the only format that reliably drove measurable profile views (8.6 per post on average). This is the format that shows up most often in the boring middle.

Template:

[Hook line, number, personal statement, or contrarian claim]

[3-5 short lines expanding the point]

[Pair with a single strong image that adds a visual layer, not a repeated caption]

[One line inviting comment or reflection]

AI prompt:

I'm writing a LinkedIn post that will be paired with one strong image.
Here's the point I want to make: [describe the idea]. Write a short,
scroll-stopping caption (under 100 words) that works alongside an image
rather than describing it. End with a line that invites a comment.

5. The contrarian myth-correction

Why it works: Posts naming and challenging a common belief (”Replying to connection requests is a waste of time”) averaged 3,836 impressions with a strong engagement-per-impression ratio.

My top contrarian post reached 79k impressions with measurable profile-view lift, solidly boring-middle territory, and solidly repeatable.

Use my post format from this post as a template for yourself and let me know how it performs

Template:

"[A belief your audience holds]."

That's what [someone/most people] think. It's wrong, or at least incomplete.

Here's what actually works:
[Your alternative view, 2-3 points]

[Why it matters / what changes if the reader adopts this view]

AI prompt:

Here is a common belief in [your industry]: [state the belief]. Write a
LinkedIn post that opens by stating this belief plainly, then challenges
it with a specific, defensible alternative. Use short paragraphs, no
jargon, and end with a practical implication, not just an opinion.

The weekly system, not a one-off post

None of these five formats produce results in isolation. The 607% year-on-year growth came from applying a mix of them consistently over many months, not from any single post. A realistic starting cadence:

Add a contrarian myth-correction post roughly every two weeks to maintain differentiation from competitors posting generic advice.

If you’re starting from zero

None of this requires an existing audience. The boring middle isn’t about scale, 590 of my posts sat there regardless of how big my following was on the day I posted them.

What matters is the format, not the follower count behind it.

If you’re posting to 200 connections instead of 40,000, the same five formats will still separate your forgettable posts from your repeatable ones.

The absolute numbers will be smaller. The pattern won’t be.

Fix your profile BEFORE you chase impressions

My headline says who I help and the result they get from working with me. My featured section sells. If your profile doesn’t name the problem you solve, every profile visit earned is wasted. Fix that first

Time Your Posts for Maximum Visibility

I’ve observed that posting when more readers are online boosts impressions and engagement.Typically, weekday early morning 7-9am works well as well as on weekends where there is less competition on the platform.

Many of my viral posts went live around 7.15-7.30 BST but what will work for you is when your audience is online.

So, plan your most promising posts for high-traffic windows. If you’re not sure when those are, experiment and watch engagement patterns.

Pro tip: Pace your posts. No more than onec a day unles there is a specific reason to ensure  each post gets the attention it deserves.

Engage

This is a critical strategy to make your LinkedIn posts seen by more people.

High engagement often comes when others repost or comment on your post, exposing it to their followers and connections. To encourage that, you should be an active participant on the platform.

Spend time interacting with other members’ posts: leave genuine comments, repost posts you love (with a thoughtful remark) and build relationships.

This raises your profile. So when you do publish a great post, those people are more likely to see it and boost it.

In short, be visible to get visible: the more you support others, the more they’ll support your work, helping your great posts go viral.

What this data cannot tell you

Impressions, engagement and profile views are leading indicators, not proof of downstream business outcomes such as newsletter subscribers, booked calls or new clients, this dataset doesn’t capture off-platform actions like DMs or sales conversations.

What transfers reliably across accounts is the mechanism, not the specific numbers: concrete hooks outperform vague ones, single strong images outperform generic ones, and consistency compounds reach in a way that no single post can replicate.

Bottom line

46.6% of my impressions came from 590 ordinary posts built on five repeatable formats.

If you’re chasing your next viral post, you’re optimising for the smaller, less controllable half of this equation. The boring middle is where the actual system lives, and it’s the only part of this you can build on purpose.

P.S. You do not have to work out LinkedIn out alone. If you would rather have hands-on, bespoke support building your strategy and content, take a look at my LinkedIn consultancy packages here

Connect with me

Substack | LinkedIn | YouTube | Website | Facebook | Instagram

TL;DR

• published a finding that writing in the LinkedIn app beats copy-pasting from ChatGPT or Claude by 22% reach. It was wrong, and he has retracted it.

• His readers challenged the mechanism, he answered “don’t know yet” to a direct question, LinkedIn’s creator product lead replied “no it’s not”, and the correction went out inside a week.

• LinkedIn published its own algorithm research on 12 March 2026. It is free and detailed, and section 9 sets out what it says about reach and engagement.

• No official LinkedIn source gives a percentage reach penalty for anything. Not links, not AI, not any format.

• Go after the practice. Leave the person alone.

I am posting this out of my usual order, because there has been a row on LinkedIn all week and I think most of it has been aimed at the wrong target.

Within a week, one wrong finding was questioned in his own comments, answered honestly, contradicted on the record by a LinkedIn product leader, updated in the original post, and then retracted in his newsletter under the heading “I Was Wrong About the -22%”. That is the correction system working, in public, faster than most newspapers manage.

Three arguments follow. The people going after him have picked the wrong villain. The way the error got caught should worry you more than the error did and the document that answers most of what everyone was shouting about has been free and public since 12 March 2026.

Subscribe now

1. The claim, and what happened to it

Point three of his post read: “This one is brutal. Writing your post directly in the LinkedIn app increases reach by 22% compared to copy-pasting from ChatGPT or Claude. LinkedIn apparently wants to see you type. I don’t love this one, but the data is the data.”

The post drew 285 comments. A large number of them went straight at that one finding, and they went at the mechanism rather than the man.

Darren Gibb put the objection most cleanly: “I can understand it ‘seeing’ if you’ve copied and pasted, but how does it decide between from claude/chatgpt and word? It’s a clipboard it’s being pasted from either way.”

Nick Mattar raised the confounders: “Maybe shorter posts are getting better results? Or could it be that posts written directly on the site are usually written on desktop and not in the mobile app?”

Kate Chidzik named the cost to anyone working under compliance: “especially when working with compliance where everything has to be approved way in advance, and then to eliminate the possibility of typos etc we copy and paste exactly what’s been approved.”

I asked him myself, in the thread, whether it meant scheduling tools reduce reach.

His answer to me was the most revealing thing in the whole episode: “don’t know yet, as we need to find out whether it’s the typical AI structure (once copied from Claude / Chat GPT) or the pasting itself. Don’t think it’s the latter though.”

He said “don’t know yet” in public, a week before the formal correction, to a direct question. He did not bluff.

2. The moment it actually broke

Three words from the person who leads creator product at LinkedIn. For the avoidance of doubt on who that is, MediaPost quoted him in June 2026 as LinkedIn’s “lead creator, product”, announcing the new audience-reach metrics.

Van der Blom then adjusted in the same thread: “I suspect it has something to do with copying directly from Claude and Chat/GPT without editing. Official API approved tools don’t show the same results.” He added: “agree that LinkedIn should fix UX when drafting inside LinkedIn.” An UPDATE line went on to the original post. The full retraction followed in his newsletter.

3. Why this should unsettle you

Look at how that correction happened.

That is the entire quality-control mechanism for an industry that sells advice to hundreds of thousands of professionals. One product manager, in a comment box, on a Tuesday.

It did not happen because a methodology was published and audited. It happened because Eli Igra Serfaty knew which LinkedIn employee to tag, tagged him, and that employee decided to answer in public.

Every link in that chain is optional. Someone in the audience has to know who to ask. That person has to be willing to put an employee on the spot in a thread with hundreds of readers watching. And the employee has to judge that a three-word public contradiction of a well-known commentator is worth their afternoon.

Remove any one of those and the 22% enters the bloodstream. It gets quoted in decks, taught in workshops, and repeated back to me by clients for the next two years, the way several other figures with no traceable source still are.

That is the quality-control mechanism for an industry selling advice to hundreds of thousands of professionals. A well-connected reader and a willing employee. It worked this time.

4. Why one person is the wrong target

Here is what I noticed while researching this piece. Many of the people criticising him hardest have been quoting his numbers in their own posts for years. The research was excellent while it agreed with them and fraudulent the moment one line moved.

Read the actual thread and the contrast is stark. It is mostly civil, mostly curious, and full of people asking sensible questions and getting answers. The nastiness happened somewhere else, later, among people who mostly were not in it.

He answered the hard questions under his own name. He said “don’t know yet” when he did not know. He put “I Was Wrong” in a headline and sent it to his own list, calling it “a bit of wishful thinking on my part”. He then delayed his next Algorithm Update to November to do the work properly, which anyone who has ever had a product date to protect will recognise as expensive.

If that is what gets you a week of quote-posts, we are teaching everyone else to say nothing and let the wrong number stand.

I also have to be straight about my own position. I have taught from these reports. I have cited figures from them in training rooms and in this newsletter. His published description of the method amounts to a self-selected sample of active creators using a partner scheduling tool. Everyone who repeated those percentages as though they described LinkedIn as a whole was claiming more than the data could carry. That includes me.

5. The problem the correction did not fix

Now the part that has to be said, because printing only the generous half would be its own kind of dishonesty.

The corrected version in Richard’s newsletter reads: “The suppression isn’t about copy-pasting itself. It’s about origin.” Content pasted straight out of an AI tool and published as-is “can see suppression up to 25%”, while content scheduled through approved third-party apps, drafted in Word, or dictated by voice is not penalised the same way.

That may well be right. Notice what happened to the number, though. One unsourced percentage was replaced by another unsourced percentage, and the second is larger than the first. Nothing published by LinkedIn supports either.

He diagnoses this himself, earlier in the same email: “Everyone’s reaching for a single percentage to settle a question that was never that simple.”

He is right, and it applies to his replacement figure too. Treat the 25% as a working hypothesis rather than a measurement.

6. The week this belongs to

Three things happened before the row.

1. On 30 July 2026, LinkedIn added a “seems like AI slop” reporting button. LinkedIn’s chief product officer said the reports “will provide a source of signal that will allow LinkedIn to tune its AI models to better identify slop”, and the company said it is “introducing new classifiers to identify if a post is AI slop or other low-quality content”. At the same time LinkedIn retired its own “enhance your post” writing feature, replacing it with one that proofreads rather than rewrites.

2. On 12 August 2026, Substack launched AI detection through Pangram, scoring posts of 100 words or more for what percentage reads as machine-written. Professor Sam Illingworth’s objection, quoted by NPR, is the one to keep: the tool “creates an environment of suspicion, and you’re guilty until proven human.” He also asks the question no score can answer. “Imagine I had an entire idea myself, and then I ask AI to riff with me on that idea. What percentage of that is AI?”

3. Everyone is reaching for a single number to settle a question about authorship that was never numerical. The 22% was one more attempt at it.

7. Solution: There is an official guide, and it is free

On 12 March 2026, LinkedIn’s engineering blog published “Engineering the next generation of LinkedIn’s Feed”, written by Hristo Danchev. The same day, Tim Jurka, a Vice President of Engineering at LinkedIn, published a companion piece on the member-facing changes. Behind both sit two technical papers written by LinkedIn’s own engineers and posted to arXiv, the open research archive.

This is the closest thing to an official manual LinkedIn has ever released. It is not a leak, not a rumour and not a paid report. It carries LinkedIn’s name, and it costs nothing…. And most people don’t know it even exists.

I would rather you read it than read me summarising it, so every link is at the bottom.

8. What the official record says

Two stages run the feed.

1. Retrieval narrows the pool. LinkedIn fine-tuned Meta’s LLaMA 3 as a dual encoder to turn both members and posts into text-based embeddings, then retrieves roughly 2,000 candidates from hundreds of millions in under 50 milliseconds.. Two details matter to anyone writing posts. Only your positive interactions build the model’s picture of you, because the engineers found that including everything you scrolled past made it worse. And popularity is fed in as a percentile rather than a raw count, a change that improved recall by 15%.

2. Ranking then orders what survives. LinkedIn calls this the Generative Recommender, described by 24 of its engineers in a paper submitted on 12 February 2026. It reads more than a thousand of your past interactions as a sequence, and names the two actions carrying the most weight: Long Dwell, meaning time spent on a post beyond a threshold that varies by post type, and Contribution, meaning a like, a comment or a share.

   In live testing it produced a 2.10% increase in time spent and a 3.52% increase in likes, comments and reshares.

Three further points worth holding onto.

1. Popularity remains a strong signal in its own right. Removing candidate popularity features cost the model 2.5% of its Long Dwell accuracy.

2. Your profile does the heavy work when your activity is thin. Profile embeddings added more than 2% accuracy for members with fewer than ten historical actions, and they refresh daily. With a dormant audience, your headline and About section are carrying you.

3. Topic beats follower count for out-of-network reach. LinkedIn’s own wording is that members discover “content from authors you don’t follow, on topics you care about, expressed in different terminology than you might use”. Van der Blom’s second finding, about an interest graph sorting you by topic rather than by followers, points the same way as LinkedIn’s own engineering. He was reading the platform correctly even where his percentages cannot be checked.

9. What LinkedIn actually says about reach and engagement

This is the part worth your time, because it comes from LinkedIn rather than from any of us.

Engagement is not a proxy. It is the target. The ranking paper is explicit that the model predicts a set of actions and combines the scores into the ranking. Long Dwell and Contribution are the two it names. Reactions and comments are not evidence that something worked, they are the thing the system is trying to produce.

Being seen and ignored is a signal. The retrieval paper describes “hard negatives” as posts that were shown to a member and received no engagement, and says the model learns from them. Adding two per member improved recall by 3.6%. Impressions without engagement are not neutral.

What you scroll past does not shape your feed. What you engage with does. The engineers found that including everything a member was shown made the model worse. Filtering to positive engagement only made it learn “much more effectively”. Curating your own feed by engaging deliberately does more than hiding things.

Popularity still counts. Removing candidate popularity features cost the model 2.5% of its Long Dwell accuracy, and popularity is fed in as a percentile so that a post in the 71st percentile of views is understood as such. Early traction on a post is doing real work.

Your profile is a ranking input, not a shop window. Profile embeddings added more than 2% accuracy for members with fewer than ten recent actions, and they refresh daily. For anyone with a thin posting history, and that is most people in a professional firm, the headline and About section are the strongest lever available.

Reach outside your network is sorted by topic. LinkedIn’s own description is discovery of “content from authors you don’t follow, on topics you care about, expressed in different terminology than you might use”.

Speed matters more than it used to. Retrieval runs against a live index in under 50 milliseconds, and LinkedIn says that when news breaks, relevant posts appear “within minutes, not hours”.

The Help Centre names three signal families, and the wording is worth reading literally. Identity covers your profile, workplace and skills. Content covers how often a post is viewed or engaged with, what it is about, how recent it is, and “how constructive or professional the conversation is”. Activity covers what you reacted to, who you interact with, and, in LinkedIn’s own phrase, “content you spend the most time viewing”. That last one is dwell time, written for members.

Demographics are not a signal. LinkedIn stated in November 2025 that its systems “do not use demographic information (such as age, race, or gender) as a signal to determine the visibility of content”, and that it runs fairness tests before model launches.

There is more competition than there was. The same November 2025 post notes that “the volume of content created daily on LinkedIn has grown rapidly over the past year, which means more competition for attention”.

10. What LinkedIn says it is reducing, and what it has never said

Tim Jurka’s March 2026 piece is the only place LinkedIn has been specific about demotion. Under a heading about generic content and engagement bait, he names posts along the lines of “Comment ‘Yes’ if you agree”, posts pairing a video with unrelated text, which “should no longer gain additional reach”, and recycled thought leadership carrying no substance.

He states that LinkedIn is “working to make engagement pods ineffective and curb comment automation, including the use of third party software or browser extensions, which are not allowed on LinkedIn”.

Now the part I would print out.

Across LinkedIn’s engineering blog, its Help Centre, its policy pages and its published research, I could not find a single official statement giving a percentage reach penalty for anything. Not for links, not for AI-assisted writing, not for any format. There is no official guidance on hashtags, posting frequency, ideal post length, best times to post, or a golden hour. The Help Centre page about content distribution that gets quoted constantly was last updated roughly three years ago.

11. The withdrawn paper nobody mentions

In January 2025, LinkedIn researchers posted a paper about 360Brew, a 150 billion parameter model, describing it in its own words as a “research pre-production model”. Half the LinkedIn advice industry then started telling people that 360Brew runs the feed.

All four versions have since been withdrawn. The file returns a 404 error, and the archive’s administrators noted it was removed because the submitter did not have the right to agree to the licence at the time of submission. LinkedIn’s March 2026 description of its live feed does not mention 360Brew anywhere.

I am not writing that from a position of superiority.

360Brew has appeared in my own material. I took it from the same secondary sources everyone else did and never went back to check whether the primary document still stood. That is the same failure this week has been about, and a better use of the argument than deciding one Dutchman is the villain.

12. A test you can apply to any report, including mine

1. Who was in the sample, and did they choose to be in it? A dataset drawn from the paying users of a scheduling tool describes ambitious creators, not the platform.

2. Does the report state a sampling frame, a margin of error, or any limitation? If the marketing says “no guesswork, pure data”, treat that as a claim rather than a method.

3. Can the specific number be traced to a primary source, or only to another commentator quoting it?

4. Is the finding mechanically plausible given what the platform has published about its own system? Point three failed this test in the comments before it failed anywhere else.

5. When a figure changes between editions, is there a changelog?

Apply all five to this newsletter as well. I would rather you did.

The takeaway. The correction worked, and it worked because someone with access happened to be reading. That is too thin a thread to hang an industry’s evidence base on. LinkedIn has now published enough of its own machinery that you can check most claims against the source in an afternoon, for free, and most of the people arguing this week were arguing about a document they had not opened.

Go after the practice. Leave the person alone. He answered the questions, said “don’t know yet” when he did not know, and printed his own correction. That is more than most of us manage.

Leave a comment

This week I went back through every setting on the screen, and every number on the analytics tab. Some of what I found surprised me. Two things in there change how any of this actually pays off, and neither of them shows up in the setup guides.

Every label, limit and screenshot below came off my own screen on 25 August 2026. Not from a blog. Not from a guide. From the newsletter I have been publishing for years.

What has changed…

LinkedIn retired the Creator Mode toggle in 2025 and then newsletters, deeper post metrics and profile links moved into the standard dashboard for everyone. LinkedIn Help now states plainly that “All LinkedIn members have access to create a newsletter on LinkedIn” . Your accounts still have to qualify, needing more than 150 followers or connections, a record of original content and an account in good standing.

I will come to it after the setup.

The setup, step by step

Do this on desktop. The article editor is better there and some of what follows is hard to reach on mobile.

Step one: find the create option

From your LinkedIn homepage, click Write article in the share box at the top of the feed. It sits next to Video and Photo. That takes you to the article editor.

In the editor, click the Manage dropdown near the upper right. You will see eight items: Drafts, Scheduled, Published, New draft, Newsletters, Create newsletter, Help, and Give feedback.

Click Create newsletter.

If you have gone looking for this in your analytics dashboard and found nothing, that is why. The old creator tools section on the analytics page has gone, and Newsletters and Lives now sit in the bottom-left area of the homepage instead (Cindy Dodd, July 2026). Going through Write article is the more reliable route.

Step two: the title, and its 30-character limit

The dialog is headed Newsletter details. The first field is Newsletter title, and it is required.

Underneath it, the screen says: “30 maximum characters allowed.”

Thirty characters is tight. “The Practical Finance Letter” is 28. There is no live counter that ticks up as you type, only that static line of helper text, so count before you commit.

LinkedIn’s own advice is to choose a name describing a clear theme your audience can understand, and its Help pages tell you to avoid emojis in the title (LinkedIn Help). Name it around the outcome your reader wants rather than around yourself. “Weekly Thoughts” and “Sarah’s Newsletter” tell a reader nothing and waste half your characters.

One thing worth knowing before you type: once the newsletter exists and the title field is populated, the dialog shows a warning reading “Changing the title may confuse your subscribers.” You can change it. LinkedIn would rather you did not. Get it close to right now.

Step three: the description, and its 120-character limit

The Newsletter description field is also required, and the limit shown on screen is “120 maximum characters allowed.”

Below the field, LinkedIn tells you where it goes: “This description appears alongside your newsletter title.”

One hundred and twenty characters is roughly one sentence. Use it to say what the reader will be able to do, not what the newsletter is about. If your title is a name rather than a promise, the description has to carry the promise on its own.

Mine reads: “The newsletter giving you the power to effectively and practically leverage LinkedIn in the palm of your inbox.” That is 110 characters. Ten spare, and a word in there I would not use today.

Step four: the frequency, and the fourth option

How often do you want to publish? is a required dropdown. It opens on “Select one” and offers exactly four choices: Daily, Weekly, Biweekly, Monthly.

Daily surprises people. Most guides list three options. There are four.

Pick the one you can survive rather than the one that sounds committed. Weekly buys more visibility at the cost of your Sundays. Biweekly is the sustainable answer for most people with a job.

Step five: the logo, at 300 by 300

The image control reads: “Add an image or logo for your newsletter to increase engagement.” The button says Upload image. The on-screen guidance is: “The recommended image size is 300x300 pixels.”

This is a logo, square, shown on your newsletter page and in the subscriber dashboard. It is not the picture at the top of each edition. Those are two different images and mixing them up is common.

LinkedIn’s best practice page confirms both: a logo at 300 by 300 pixels for the newsletter, and a cover photo at 1920 by 1080 pixels for each individual article inside it.

The control is optional. Fill it in anyway. A newsletter with no logo looks abandoned before it has started.

Then click Done.

Step six: the cover image or video for each edition

Back in the article editor, above the title field, there is a box reading “Add a cover image or video to your article”, with a single button: Upload from computer.

That is the entire instruction. There is no tooltip, no info icon and no stated file type, dimensions, aspect ratio, duration or file size anywhere near that control. I looked, including in the underlying page code. LinkedIn’s best practice page gives 1920 by 1080 pixels for an article cover photo and says nothing at all about video (LinkedIn Help).

For video covers, then, work to sensible limits and expect to be told off by the upload screen rather than in advance. Fifteen to thirty seconds, filmed wide rather than upright at 1920 by 1080, you talking to camera, saying what this edition covers and who it is for. Not a trailer. Not branded motion graphics. A face and a sentence. The notification is where most newsletters are won or lost, and a face in it beats a stock photograph.

Step seven: publish, and the box that appears afterwards

Write the article. Confirm your newsletter title is showing near the upper right, which tells you it will publish into the newsletter rather than as a standalone article. If it is not showing, go back to the Manage dropdown and select the newsletter. Then click Next, then Publish.

A window then pops up asking for more context. LinkedIn describes it as a place to comment on what this edition is about, ask your readers questions, and use the @ symbol to mention people or Pages. When you click Post, the edition is also shared to the feed.

Most people rush this box or leave it empty, so the feed post reads like a link drop. Given what the numbers below show about where reading actually happens, that box deserves three or four proper lines making the case for reading.

Step eight: the invitation only fires when you publish

This is the step that decides whether your launch works, and it is the one people get wrong by not knowing it exists.

LinkedIn Help states: “By default, your connections and followers will be invited by notification to subscribe to your newsletter once the first article of your newsletter has been successfully published”

Creating the newsletter sends nothing. Publishing the first edition sends the invitation to your entire connection and follower base. You get one of those.

Do not create the newsletter and publish a thin welcome post to test the plumbing. Whatever you publish first is what your whole network is invited on the strength of.

Step nine: changing things later

You can edit everything. Go to your newsletter page, click Edit below your name, and the same dialog reopens with the title, description, frequency and image all editable.

A limits note, and a genuine contradiction. LinkedIn’s Help pages say you can create up to five newsletters and unlimited editions within each, with a two-week wait after creating or deleting one before you can create another. The text inside the edit dialog on my own screen says something different: “Limit 1 newsletter per member.” I have two. Treat the five as the real figure and the dialogue text as out of date, and do not build a plan that depends on either number until you have seen your own screen!

How to find the settings panel

Open any published edition of your newsletter and click Edit article. In the editor, look for the Manage dropdown in the top-right of the toolbar and click Settings. A dialog headed Settings opens, with a section labelled SEO and two fields underneath.

A few older guides describe this panel as a gear icon in the top-right. On my screen on it lives inside the Manage dropdown. If your build differs, look in both places.

The four-step path to the SEO settings panel: open a published edition, click Edit article, click Manage, click Settings.

Field one: SEO title

The field is labelled Title. The helper text on the screen reads verbatim: “We’ll use your added SEO title in place of your article title for search engine result pages, such as Google search.”

There is no visible character counter next to the field, unlike the description below it. The field is hard-capped at 60 characters. If you try to paste in more, it truncates without warning. LinkedIn Help confirms the 60-character limit.

This is not the title your subscribers see on LinkedIn. It is the one Google shows the world when someone searches for what your article is about. Sixty characters, front-load the keyword.

“How to write a LinkedIn newsletter that ranks on Google” rather than “The one thing nobody tells you about newsletters”.

Field two: SEO description

The field is labelled Description. The helper text reads: “We’ll use the SEO description in place of the first few lines of your article on search engine result pages. We suggest utilising keywords, summarising your writing, and aiming to write between 140-160 characters.”

Underneath the field a counter runs from 0/160, and static helper text confirms “160 maximum characters allowed.”

This is the little grey text under the blue link on the Google results page. Without it, Google decides for itself, usually pulling the first line of your article. Which is fine if your first line sells the piece and unhelpful if it opens with “Welcome back everyone, hope you had a lovely weekend”.

What is not in this dialog

Some older guides mention a third field, a custom URL slug you can rewrite for each edition. On my screen on 25 August 2026, there is no such field. The Description field ends, the counter appears, and immediately below that are the Cancel and Save buttons. Two fields, not three.

Moving on to the more useful point: you can add these two fields retrospectively to every edition you have already published (Marisa Shadrick, March 2026). Every edition in your archive is a piece of content Google is indexing right now. If you spent an evening going through your back catalogue and filling in the two fields per edition, you would be doing more for your search visibility than most people who pay agencies to do it.

Do it once for every past edition. Do it every time you publish a new one. This is the setting that changes everything, because it turns each edition from a piece of content with a two-day notification life into a piece of content that keeps finding readers for years.

The email does not reach your subscribers

Now the part that changes how you should think about all of the above.

LinkedIn Help says publishing an edition sends push, in-app and email notifications to all your subscribers. Most people read that as everyone.

The Link Tank shows 13,097 subscribers. Here are the three most recent editions as of 25 August 2026, with the figures LinkedIn labels Email sends and Email open rate.

The edition published on 20 August recorded 7,696 email sends against those 13,097 subscribers, at a 25 per cent open rate.

The 14 August edition: 7,684 sends, 26 per cent. The 12 August edition: 7,677 sends, 27 per cent.

Divide sends by subscribers and each one comes out at roughly 59 per cent.

Around four in ten subscribers were not emailed at all, and the proportion held almost perfectly steady across three editions.

I am not the only one seeing this. Toni Hunter reported sends against subscribers running at 64.5 per cent for a recent edition of her Finance Friday newsletter to 3,400 subscribers.

Treat somewhere between 35 and 60 per cent as the working range. I have found no published LinkedIn guidance explaining what determines it, so this is an observed pattern rather than a documented rule. Check your own before you trust anyone’s figure, mine included.

Two things follow.

1. Your subscriber count is not your reach. Thirteen thousand subscribers is not thirteen thousand inboxes. Planning as though it is will make every edition feel like a failure.

2. The on-platform read is not the consolation prize. That 20 August edition recorded 2,779 article views. A 25 per cent open rate on 7,696 sends is roughly 1,924 opens. More people read it on LinkedIn than opened the email. If you are writing for the inbox and treating the feed appearance as an afterthought, you have the priority backwards.

The mechanic that actually compounds

Here is where growth comes from and it is not the email.

LinkedIn Help states: “When a member starts following another member or a page with newsletter, a notification inviting them to subscribe to the newsletter is automatically sent”

The invitation is not a subscription. The follower still has to accept it. What makes it valuable is that it fires automatically on every new follower, forever, without you asking anyone for anything.

Ordinary posting builds followers. Followers meet the subscription invitation. Some accept. Every edition is another chance at the ones who did not.

That is also why the first three editions carry more weight than the ten after them. Someone who followed you last week is deciding whether to accept that invitation based on whatever you have published recently.

One caution before you build a business on it. You do not own the list. LinkedIn does not expose subscriber emails to the creator. You cannot export them. You cannot email them from another tool. If LinkedIn removes your newsletter, everyone is gone.

A LinkedIn newsletter is a rented asset with excellent distribution, which is an argument for running one and also an argument for not running only one.

This is why you need a Substack newsletter as your ownership arm.

The first three editions

Three editions, three different jobs, in this order.

The reason three is the number: a reader who sees edition one and does not subscribe has to be caught again by edition two. Someone who subscribes at edition one and finds edition two thin will mute the notifications, and a muted subscriber is worse than no subscriber, because the count still looks healthy while the open rate rots.

1. Edition one is the reason it exists. One specific problem your reader has, one specific answer, delivered in full. Around 800 to 1,200 words. The common mistake is writing edition one about the newsletter itself. Nobody subscribes to a table of contents. Write the first useful thing and put the promise in one line at the top.

2. Edition two is the proof. The reader who liked edition one is checking whether you can do it twice. One worked example taken all the way through, with the awkward part left in rather than edited out.

3. Edition three is the framework. Something the reader will use again without you. A named checklist or test they can remember. This is the edition that converts a casual reader into a long-term subscriber, because a framework implies there are more where that came from. Around 1,200 to 1,800 words, and worth the most work of the three.

Reading your analytics

The labels differ depending on where you look, which catches people out.

On the newsletter analytics page you get Article views, New Subscribers, Impressions and Engagements, with a seven-day chart and a note that daily data is recorded in UTC. Mine currently reads 3,858 article views and 36 new subscribers for the past seven days.

The email figures are not there. Email sends and Email open rate appear only on each individual edition’s post analytics, under a section headed Article performance, alongside Article views. To work out your sends ratio you therefore have to open an individual edition rather than the newsletter dashboard.

Three comparisons tell you whether this is working.

Sends divided by subscribers. Almost nobody checks this. If it is well under a third, your list is bigger on paper than in practice and the feed side needs the attention.

Article views against estimated opens. Multiply sends by open rate to get rough opens, then compare with article views. If views win, your feed post is doing the heavy lifting and deserves more care than your email subject line.

Followers gained per edition. LinkedIn does not attribute new subscribers to a specific edition. The closest per-edition figure is “Followers gained from this post”, which for my last three editions was 2, 2 and 1. Small numbers, so read the trend across several editions rather than any single one.

Where people go wrong

Publishing edition one just to get started. That first publish spends your one network-wide invitation. Make it your best work.

Mass-inviting connections. LinkedIn lets you invite connections, and the sensible approach is targeting relevant ones and spreading invitations over time Mass invitations get you subscribers who never open, which poisons your sends ratio from day one.

Leaving the SEO fields blank. Every edition without an SEO title and SEO description is being served to Google with whatever LinkedIn chose to auto-generate. Go back and fill them in. Even for the ones you published years ago.

Leaving the context box empty. The feed post may outperform the email. Write it.

Declaring a cadence you cannot keep. My own setting says Monthly while I published ten editions between 26 June and 14 August. Over-delivering against a modest promise costs very little. Declaring Weekly and then publishing nothing for five weeks trains people to stop expecting you, and that is the version that does damage. I am changing my setting to match what I actually do.

Building only here. You cannot export the list. Run this for the distribution and keep something you own alongside it.

The takeaway

Access is no longer the constraint. Anyone can publish a LinkedIn newsletter now, and the whole setup takes about ten minutes once you know the title gets 30 characters, the description gets 120, and the invitation fires the moment you publish rather than the moment you create.

Then there is the setting that changes everything: the two SEO fields inside each edition. Manage, Settings, SEO title, SEO description. Filled in properly, they turn a LinkedIn newsletter from a notification you send once into a piece of writing Google can find for years.

Set the four setup fields properly. Save your best work for edition one. Then open a published edition, find Email sends, and divide it by your subscriber count. Then fill in the two SEO fields, on that edition and every one before it.

Have you filled in the SEO settings on your LinkedIn newsletter, and did you know they were there?

Leave a comment

Hi I’m Melanie, a LinkedIn consultant & strategist who turns insider knowledge and sharp strategy into measurable results for ambitious professionals.

If you’ve ever written a candid post, hovered over the publish button, and then closed the tab, this one is for you.

The difference between a candid post that makes a career and one that ends it has almost nothing to do with how brave the writer was. It has almost everything to do with one question the writer never quite asked.

Here’s the question and nine others that go with it.

Subscribe now

The post I nearly didn’t publish

Two months ago I turned fifty. I wrote a short post about it on the morning of my birthday. I nearly deleted it three times before I hit publish.

It has since garnered 206,316 impressions, 1,400+ reactions and 441 comments. Not because it was clever. Because it named something a lot of people my age are quietly carrying about milestone birthdays and were relieved to see somebody say out loud.

The birthday post. 206,316 impressions, 441 comments, three near-deletions.

Every LinkedIn coach tells you to be vulnerable. The data tells a different story.

A separate sentiment analysis of 4.6 million LinkedIn posts went further. Emotional tone had almost no effect on virality on its own. What mattered was subjectivity. First-person, opinionated writing outperformed detached commentary by nearly seven to one, regardless of whether the mood was hopeful or bleak.

The working principle: write something only you could write, from a position only you occupy, about something a lot of people are already carrying.

That is what the birthday post did.

It is also what Emily Lyons’s 26.4 million-impression post did about what women learn to hide when they succeed.

It is what Jonathan Frostick’s 2021 post did, the HSBC manager who had a heart attack at his desk and opened with “I needed to meet my manager tomorrow, this isn’t convenient.” His post received 298,000 reactions. Not because it was about him. Because it was about the millions of people whose first thought during a health scare is still a work deadline.

The cautionary tale

Now the counter-example.

Braden Wallake’s August 2022 post, as syndicated by CNBC. The original was later deleted.

In August 2022, Braden Wallake, a marketing agency chief executive, posted a tearful selfie of himself after laying off staff. The opening line: “This will be the most vulnerable thing I’ll ever share.”

Journalists called it “narcissistic virtue signalling of the highest order” and “tone-deaf” (The Journal, 2022). Four years on, Wallake is still known as “The Crying CEO”.

Not ideal.

The mistake keeps being made…

In June 2024, Matthew Baltzell, chief executive of Cap X Media, posted a detailed account of firing his first employee, describing the “short and direct” ten-minute meeting. X users seized on it with “imagine getting fired, heading over to LinkedIn, and seeing this.”

My birthday post and Wallake’s layoff post both opened with a version of “I almost didn’t share this.” Both went viral. Only one is remembered as a career highlight.

The difference is where the reader sits in the story. In my post, the reader is the subject: anyone dreading a big number. In Wallake’s, the reader is the audience for the writer’s feelings.

Recognition points outward. Performance points inward. Every question below is a way of asking which one you are doing.

The specificity difference

Before the checklist, one more of my own for context.

Earlier this year I wrote about a client situation I kept seeing: employers who lose valuable staff because they refuse a small pay rise. The post opened: “It cost them £50,000 to avoid a £10,000 raise.”

The £50,000 post. 145,600 impressions, 426 reactions, 79 comments, 18 reposts. Written in about twenty minutes.

What made it work was the arithmetic. Every number in the post is specific: £60k salary, £70k ask, £40k pipeline lost, £10k in recruiter fees, three weeks onboarding.

Vague version: “A friend told me about a company that lost a great employee because they wouldn’t give her a raise.” Nobody shares that. The specific version puts a mirror in front of every leader who has ever made the same call.

The recognition happens through the detail, not through the emotion. Which brings us to the checklist.

Why LinkedIn amplifies both kinds

LinkedIn’s algorithm rewards dwell time and comments. Both recognition posts and performance posts generate both. A recognition post gets long comments from people saying “yes, this is my life”. A performance post gets long comments from people saying “this is exactly what’s wrong with LinkedIn”. The platform, indifferent to the difference, boosts both.

The judgement has to come from you before you press publish. LinkedIn will not save you from a bad post. It will happily accelerate it.

The ten questions to ask before you post

Each takes about thirty seconds. If any answer makes you uneasy, hold the post back for twenty-four hours. The next morning, you will know.

1. The recognition test

Does this post give language to something my reader is already carrying, or does it perform something at them?

Strip yourself out of the post entirely. If there is still a takeaway for the reader, publish. If the only reason it is interesting is that it happened to you, you are performing.

2. The Monday morning test

Would I stand by every line of this on Monday morning, sober, in front of the person whose opinion of me matters most?

Drafts written at 11pm on a Sunday go wrong because they were written in a state you will not be in when the responses arrive on Tuesday.

3. The named parties test

If I name someone, or describe them so specifically they can be identified, could they use this against me?

The bar is could, not would. If the answer is yes, strip the detail or don’t post.

4. The specificity test

Have I given the reader one specific number, one specific moment, or one specific line of dialogue?

“I had a hard year” is empty. “I lost three clients in April, sat in my car for an hour before a school pick-up, and didn’t tell my husband for a week” is a story. The £50,000 post is this test in action. One specific detail makes the reader feel seen.

5. The lesson test

What will the reader take with them after they close the tab?

If the only takeaway is how the writer felt, the reader has nothing to use. The best truth-telling posts give the reader a sentence they can quote to a colleague tomorrow.

6. The subject-object test

Is the reader the subject of this story, or the audience for my feelings?

The deepest version of the recognition test, and the one most people fail without noticing. If the reader is the audience, hold back.

7. The 24-hour test

Has enough time passed between the thing that happened and my writing about it?

Emily Lyons hesitated before publishing. Braden Wallake did not. The difference shows in every line. When something has just happened, the post you would write today is almost never the post you should publish.

8. The consequence test

Could this cost me a client, a partner, or a role I might want in two years?

Recruiter Shelley Crane, in Fortune’s guide to layoff posts, warns that once posted, “those words are out in the world.” Two years is the right horizon: long enough that today’s urgency looks different, short enough that professional consequences are still real.

9. The “why me, why now” test

Am I the right person to say this, and is this the right moment?

Riding a trending topic with a personal story is one of the most reliable ways to look opportunistic. If the only reason your post exists is that everyone else is posting about the same thing, wait.

10. The performance test

Am I posting this because it’s true, or because I hope it will go viral?

Not every candid draft belongs on LinkedIn. Some belong in your diary. Some belong in a private message to one person. Some belong nowhere at all.

The line I come back to whenever a client asks whether to post something raw:

Would you rather be the person who wrote the post, or the person people read?

I nearly deleted the birthday post three times. I am glad I did not. Braden Wallake did not delete his either. Four years on, the difference is still visible on both our profiles ;)

Before you write your next candid post

Paid subscribers only.

I run every candid post I write through these ten questions before I hit publish. I have built the framework into a Claude skill (can also use with Chat GPT etc)) so you can do the same, without keeping the checklist open in a tab.

Paste your draft, and the skill audits it against all ten questions, marks each as Pass, Uneasy or Fail with a quoted line from your post as evidence, gives you a one-line verdict, and gives you the three edits and rewritten version that would make the most difference to your post to give it a higher reach - more impressions.

It takes about ninety seconds per draft.

Download and use skill below:

Read more

I started this by fact-checking a stat I kept seeing quoted. Someone would drop into a comment thread and say “there are 184,000 newsletters on LinkedIn and only 30% publish monthly”, or “98% of the top newsletters are owned by individuals”. I went to check. Neither is true today and the leaderboard that supposedly proves it is 18 to 30 months out of date.

I did the work properly:

I opened the biggest circulating lists, checked every one against the live LinkedIn newsletter and edition pages where the current subscriber count is visible (like this one for In the Loop), and reordered the leaderboard on the numbers I could see with my own eyes on 19 August 2026. Then I pulled the analytics on my own LinkedIn newsletter, The Link Tank, to see what the numbers actually do at a normal scale. That last part turned out to matter more than any of the lists.

1. The numbers everyone quotes are two years old

The list you are most likely to have seen is Tomislav Horvat’s LinkedIn Newsletter Statistics, republished endlessly on other blogs. It puts In the Loop at 7.4 million subscribers and Gates Notes at 2.3 million. When I opened those two pages this morning, In the Loop showed 11,534,190 and Gates Notes showed 6,523,720. The reality is somewhere between two and three times larger than the numbers being cited.

The Reletter top-50 LinkedIn chart is fresher, but its methodology mixes creator newsletters with company Pages inconsistently, and its snapshot dates lag by months. If you cite either without checking the live pages, you are quoting the past.

What to do instead: open the newsletter page directly, look for the “N subscribers” line at the top, and use that. It is the only number LinkedIn shows publicly and the only one that updates in real time.

2. The current top 15, in order

Here is the leaderboard as of this morning. I checked every entry on its live LinkedIn page. Owner type is my classification: I have counted media brands (HBR, Forbes, TED, Economist, WSJ) separately from operating companies (Microsoft, Google, IBM), because they behave differently.

3. Most of the top of the leaderboard is corporate

I have seen the claim “489 of the top 500 LinkedIn newsletters are owned by individuals” quoted as if it were physics🤪 It is not true at the top of the list.

Of the 15 newsletters above, 11 belong to companies or media brands and 4 to individuals (Bill Gates, Richard Branson, and two named columnists further down). Widen to the top 50 and the corporate share stays above 70%.

There is one obvious reason: these are the accounts with existing audiences to convert. LinkedIn News alone has 34.1 million followers behind In the Loop, and the newsletter converts a fraction of them to subscribers. An individual creator has to earn every subscriber from scratch.

4. Every subscriber does not get every edition

This is the finding that changes how you should think about the whole product. LinkedIn Help says publishing an edition sends “a notification and an email to your subscribers”. Most creators read that as “everyone gets emailed”. The evidence says otherwise.

When Gina Riley checked her own analytics, she found only around 35% of her subscribers had been emailed for a given edition (3,300 to 3,600 email sends per issue). LinkedIn support reportedly told her that sends are filtered by relevance. Nobody outside LinkedIn knows the exact rules, and other practitioners cite very different numbers, which is exactly the point: this is opaque.

My own numbers say the same thing at a different scale. Here is the analytics screen for one recent edition of my newsletter, The Link Tank:

The chart below is the same data as a funnel. Two things stand out: roughly 4 in 10 subscribers were not emailed at all, and more people read the edition on LinkedIn than opened the email (2,852 article views vs an estimated 1,997 email opens at a 26% open rate).

The takeaway: the “subscriber count” is not a mailing list. It is closer to a follow graph. If you want your edition read, you still need it to work as a post on the LinkedIn feed and as a searchable article.

5. Monthly and weekly do most of the work

I logged the declared cadence for every newsletter in the top 58. Monthly leads (26), weekly is close behind (21), biweekly (6), daily (4).

One “daily” newsletter is in fact seven weeks behind, so the declared and delivered cadences do not always match.

If you are choosing a cadence, the top of the leaderboard is telling you two things. Weekly is the sweet spot for media brands and topic-led shows. Monthly is where operators, executives and busy specialists live. Biweekly is the honest choice if you know you cannot sustain weekly.

Rule of thumb: pick the cadence you can sustain for 12 months on a bad month, not the cadence you can sustain on your best month.

6. Titles are short, plain, and topic-first

I ran the numbers on the titles of the top 58 newsletters. The average title is 2.93 words. Three quarters are 1 to 3 words long. Only three include the author’s name (Gates Notes, Ask Richard, and one columnist). Twenty-four include a company or brand name.

Search still runs on LinkedIn, and Google indexes public /pulse/ URLs within 24 to 48 hours. A title that says what the thing is beats a title that is clever about it.

If you are titling yours: two to four words, topic before personality, no colons if you can avoid them. “LinkedIn Training Weekly” beats “The Melanie Files: Real Talk on LinkedIn”. Save the personality for the deck 😁

7. The subscribe prompt does most of the growth

When someone follows you on LinkedIn, they are shown an invitation to subscribe to your newsletter. The invitation is not a subscription: the follower still has to accept.

The mechanic matters, because it is automatic and it fires on every new follower. On the Creator Science podcast, describes it as LinkedIn “actively promoting a creator’s newsletter to new followers”, and says newsletter editions now outperform regular posts on reach, engagement and conversion, while standalone articles perform at roughly 0.4x normal reach.

The practical consequence: growth is coupled to how well your normal posting works. Feed posts add followers, followers get the subscribe invitation, some subscribe, and each new edition is another chance to convert them. No single “viral” edition is required. A modest weekly increase in followers compounds into a much larger subscriber base over a year.

8. LinkedIn Help understates the mechanics

9. The create flow is deceptively simple

The panel for creating a newsletter has four fields and one dropdown. Title, description, cover image, cadence. That is it. Every strategic decision has to be made before you ever see this screen.

The three decisions the panel does not force you to make properly are the ones that decide whether the newsletter works:

1. A niche narrow enough to be searchable. Two words is the sweet spot. Broad titles (”The Weekly”) do not surface.

2. A cadence you can hold on your worst month. LinkedIn shows relative dates (”2 months ago”) on every edition, which visibly ages a stalled newsletter.

3. A distribution loop that does not depend on the email. Assume ~35 to 60% of subscribers get the email and plan the on-platform post accordingly.

10. You do not own the list

LinkedIn does not expose subscriber emails to the creator. You cannot export your list. You cannot email them from another tool. If LinkedIn removes your newsletter, everyone is gone. Creators have discovered another sharp edge: LinkedIn has quietly demoted standalone long-form articles. They still exist, but m measures their reach at roughly 0.4x a normal post, and the platform now steers writers toward newsletters instead.

So the format still exists, but it has been demoted. It is a rusting side door, not the front entrance.

A LinkedIn newsletter is a rented asset with excellent distribution, not an owned one.

The safe pattern: build the LinkedIn newsletter for reach and search, and use the CTA in every edition to pull subscribers to an email list you actually own. Two lists, one topic.

That’s the LinkedIn-Flywheel!

Should you start one?

Here is the transparent version of the flow chart. If you can answer yes to all four, start it this month. If you cannot, do the missing step first.

1. Do you post consistently on LinkedIn already? The newsletter does not create your audience. It converts the audience your posting builds.

2. Can you write 300 to 1,500 words per edition without dying? You need 12 to 24 editions before growth compounds. That is a year of monthly, or six months of weekly.

3. Do you have a topic that fits in three words? If the title needs a colon, keep working on it.

4. Do you have somewhere the subscribers can go if LinkedIn takes it down? Even a simple contact form to a Substack or a website works. Just do not skip it.

If yes to all four: press Create newsletter, pick a cadence you can hold on your worst month and treat the first ten editions as a shakedown cruise.

The one-line version

A LinkedIn newsletter is a landing page with a distribution engine wired into it. The subscriber number on the front is a vanity metric. The subscribe prompt behind it is where the growth is.

Leave a comment

Subscribe now

Most people treat a LinkedIn event as a single announcement post. They build the event, publish one “we’re hosting a thing” update, and then wonder why fourteen people register and four turn up. The announcement is not the work. It is the smallest part of it.

An event on LinkedIn is a three-week campaign with a page at the centre and a dozen touchpoints around it. Filling the room is a function of infrastructure, sequencing, and follow-through, not the cleverness of one post. Here is the full system, in the order you should build it.

1. Build the event container on LinkedIn first

Before you promote anything, build the thing you are promoting.

1. Create a native LinkedIn Event using this link and treat the page as a small landing page rather than a calendar entry.

   

2. Give it a benefit-led title, a tight description that names who it is for, the problem it solves, and what the attendee walks away with, and a proper banner.

3. LinkedIn supports banner sizes including 480x270 or 1280x720 pixels, so size the creative rather than letting it crop badly.

Gated or ungated. That is the question

The single most important decision is gated or ungated: If you keep the event ungated, with no registration form, you reduce friction and attract as many attendees as possible, and the more attendees it attracts, the more feed reach it earns. That suits an awareness event.

For lead generation, add a registration form, and please be aware of two things first:

1. The registration form option is only available when you create the event on behalf of a Company Page and

2. You will need to provide a privacy policy link.

Once a member clicks register, they see a form pre-filled with their name and email from their LinkedIn profile, and if you add a form, members cannot view the event without submitting it.

The form can collect name, contact details, job title, company, and country, with up to six fields in total.

Gating raises the value exchange, so only gate when the content is worth handing over an email for.

If your event runs off LinkedIn on Zoom, a webinar tool, or your own platform, you can add an external event link so people can join, though LinkedIn recommends using LinkedIn Live where possible to reduce drop-off when attendees switch platforms.

For a paid event, keep the LinkedIn Event as the free RSVP and nurture hub, and send the paid transaction out to your ticketing tool.

2. Organic promotion from your profile and your Page

This is the backbone of sign-ups, and the most underused. Announce from both your personal profile and your company Page, stating the date, time, format, and the reason it matters in the first two lines. Endorsements from people in your network raise both trust and attendance, so make the event easy to reshare with a recognisable banner and a one-line promise that reads the same across every post.

Do not lead with the invitation. Warm the audience for two to three weeks first with content on the event themes. Speaker spotlights, a short problem post (”the three mistakes I keep seeing people make with this”), and teaser insights with a soft line at the end. Think of it as a micro-curriculum. Three posts that read as lessons one to three, then the live event as lesson four. A long-form article on the topic, with the registration link in the body, keeps working long after the promotion window closes, because articles surface in search for months.

Short video earns attention here. A thirty-second clip previewing a case study or answering one question, with captions and a clear call to action, gives you another route back to the page. Share the event in groups only where the topic genuinely fits the group’s purpose, and frame the post around member benefit rather than your own promotion, or it reads as spam.

3. Direct outreach and networking

Most sign-ups come from targeted invitations rather than a broadcast. Build a list of fifty to a hundred ideal attendees using LinkedIn search filters, or Sales Navigator if you have it, segmented by role, sector, and the challenge your event addresses. Treat this as your recruitment pipeline and log every touchpoint.

Warm the list before you invite. Spend two to three weeks viewing profiles and commenting substantively on their posts, so the invitation arrives from a familiar name rather than a cold one. Then send a personalised connection request, and once accepted, invite conversationally. Avoid the generic “we’re hosting an event, want to come” message. A structure that works:

Personal observation (their role, a recent post, their focus). One insight linked to the event topic. A soft introduction to the event. A low-friction offer: “Happy to send the details if that’s useful.”

Follow up with value, not pressure. An updated agenda, a new speaker, or an extra insight will move more people than “just checking you saw this.” A warm list run this way will outperform every other channel you have.

4. Event Ads and paid distribution

If your audience is large enough to justify spend, LinkedIn’s paid tools extend reach fast. You can amplify an event by boosting an organic event post or using the dedicated Event Ad format, targeting your ideal attendees by job title, industry, company size, and location. Keep the copy short and mobile-first, with one strong visual and a single call to action.

The highest-return paid move is retargeting. Run campaigns to people who engaged with your event posts, watched your video, or attended a previous event. That is where paid distribution turns into follow-up conversations and future registrations rather than one-off impressions. If you are promoting to distinct groups, split the campaign and test a different hook per segment rather than running one message for everyone.

5. Collaborations, advocacy and attendee content

Your speakers and your team are distribution channels you already own. Give them a promo pack: approved copy, the visual assets, and a suggested posting schedule. It keeps the messaging consistent and on-brand while multiplying your reach. Ask speakers to post their own angle on the topic and to invite relevant contacts directly, which carries more weight than a corporate post.

Design the event to be co-created rather than broadcast. For in-person events, a branded backdrop earns “I’m here” posts. For virtual, prompt attendees to share one takeaway. Afterwards, reshare attendee recaps and tag the people involved, so they feel seen and the conversation carries into your next event. Inviting a respected voice in your space to co-host does more for credibility and registrations than any amount of your own promotion.

6. Engagement mechanics and lesser-known features

A few features earn more than they cost:

A poll on the event topic gathers views and gives you a natural line to “we’ll go into more depth on this at the session.” A “what would you ask our speaker about X” post collects real questions you can then answer live, which both shapes your agenda and gives attendees a reason to show up. Reveal the agenda in pieces rather than all at once, with countdown posts at seven, three, and one day out, each on a different angle: the problem, the solution, the networking, the bonus resource.

One caution on the mechanics. LinkedIn has been changing how event-related content surfaces. For Live Events, it has been removing the dedicated Event feed and making event content behave like an ordinary feed post, so check the current behaviour of your event feed before you plan your posting around it. Where LinkedIn Live is available to your Page, broadcasting a session or a short pre-event interview, promoted in advance with a clear call to register for the replay, extends your reach well beyond the live audience.

7. Scheduling and planning without breaking the rules

You can run this entire campaign organically, with no paid spend, using the event page, a sequenced content calendar, personalised invitations, and consistent engagement. A twenty-one-day calendar from first announcement to last-chance reminder, with defined post types and outreach blocks, is worth building once and reusing for every event.

I built the whole thing so you do not have to start from a blank page. The LinkedIn Event Kit holds the twenty-one day promotion calendar day by day, swipe copy for every stage from the first announcement to the no-show follow-up, and seven AI prompts that draft each piece in your own voice and stay inside LinkedIn's rules.

The calendar runs in three phases. Days one to seven set up the event page and warm your audience. Days eight to fourteen open the invitations and the personalised outreach. Days fifteen to twenty-one run the countdown, the agenda reveal, and the last-chance reminder.

Each day names the post to publish and the outreach block to complete, so you are never guessing what comes next.

Access LinkedIn Event Kit

8. Measurement, consent, and follow-through

This is the part most people skip and it decides whether the event builds anything. Track the numbers that tell you what worked: connection acceptance rate, direct message response rate, post engagement, registrations, and actual attendance. Each event is a test bed for the next.

The consent rule is not optional, whatever you are hosting. By ticking the checkbox on the registration form, a member opts in to promotional or marketing communications, and organisers should not send product communications to members who did not select it.

Downloading an attendee list does not give you licence to market to everyone on it. There is also a clock on the data. Registrant data is only available to download for up to one year from the day a member registers, after which it can no longer be accessed, so export and store it properly within that window under your own data protection obligations. If a team promotes on your behalf, agree the messaging in advance so the advocacy stays on-message.

The whole system rests on one idea. An event is not a post you publish, it is a relationship you build over three weeks and continue after the room empties. Fill the calendar, warm the list, respect the consent, and the attendance follows.

Book a LinkedIn Training Session

What is the one event promotion tactic you have tried that genuinely moved attendance, and the one that did nothing?

I read every reply.

Hi I’m Melanie, a LinkedIn consultant & strategist who turns insider knowledge and sharp strategy into measurable results for ambitious professionals.

Before we get into it, a quick “hello”.

Something I want to do more of here is bring you into the process, rather than publishing at you. You are the reason The Link Tank exists, and your replies shape what I write next.

Here is a poll before you read on. One tap, no wrong answer, and you will see the results build as other readers vote. It tells me where you are actually stuck, which is how I decide what to write next.

Whatever you tapped, it tells me what you need next, and I will write towards it. Today’s piece is for everyone who chose no. 1.

There is a free feature sitting on your profile that answers exactly that problem and many otherwise well-built profiles leave it unused

Here is what brought it to mind this week…

A friend messaged me last month, frustrated. She had spent a year building a respectable LinkedIn presence. Steady posts, a tidy profile, a few hundred new connections. She had almost nothing to show for it. No enquiries, no work coming in from any of it.

I looked at her profile for about thirty seconds and found a major problem. Nowhere on it did she tell anyone she was available to hire or how they would go about it.

This is the failure at the centre of most people’s LinkedIn. They treat the profile as a CV. A record of where they have worked, tidied up and left to sit there. A CV tells people what you have done. It does not tell them you are open for work, and it does not give them a way to start the conversation. That gap is where enquiries go to die.

There is a section built into LinkedIn that closes the gap. Almost nobody turns it on. It is free, it takes about ten minutes, and it does the one thing a CV cannot. It tells the right person you are available, and it hands them a button to contact you.

It is the Services section. Here is why it works, and how to set it up so it actually earns you something.

Subscribe now

What it is, and why it is different from everything else on your profile

When you add Services to your personal profile, LinkedIn builds what it calls a Service Page. LinkedIn displays a Services module on the profile, though exact placement and labels can vary by device and interface. A panel appears near the top of your profile, above your work history, listing the services you offer.

A “Providing services” line shows under your name.

The feature is free. You do not need LinkedIn Premium, Sales Navigator, or a Company Page to use it.

See Page in Action

Who should use this

This is not a feature for everyone. If you are looking for a permanent role, focus first on making your experience, headline and Open to Work settings clear.

But if somebody can hire you - whether you are a consultant, coach, freelancer, trainer, agency owner or fractional specialist - this belongs on your profile. It gives a buyer a route to make contact at the moment they decide they need help.

What it will not do

A Services section is not a magic lead machine. It will not rescue a vague offer, compensate for a profile that gives no reason to trust you, or replace the work of building visibility and relationships.

Think of it as the conversion point, not the whole marketing system. Your content, comments, profile, recommendations and network create interest. The Services section makes it easier for that interest to become an enquiry.

Two things make it more than another box to fill in.

1. The first is contact. Someone can send you a service request or a message straight from that panel, and in many cases they can do it without being connected to you. That matters more than it sounds. Normally, a stranger who wants to hire you has to send a connection request, wait, hope you accept, and only then work up the nerve to ask. Most of them never get past the first step. The Services section removes every part of that wait and replaces it with a button.

2. The second is discovery. LinkedIn uses the services you list to surface you when members go looking for that kind of help. I cannot give you a precise number for how much extra visibility this buys you, and I would be suspicious of anyone who claims one. The mechanism is real and it costs you nothing to switch on.

Why two small words change how people behave

A profile with no Services section leaves both questions hanging. The reader has to guess your availability from your headline, assume you take on this kind of work, then decide whether it is worth the mild embarrassment of asking a stranger. Most people, faced with that much friction, simply move on. Not because they were not interested. Because you made them do the work.

“Providing services” answers both questions before anyone has to raise them. It says: this person is open, and here is exactly what they do. You have taken away the doubt that stops the enquiry before it starts. Far more work is lost in that doubt than in any lack of demand.

Setting it up, in about ten minutes

Go to your profile. Depending on your layout, start from Open to, Add profile section, or the Services prompt in your intro area and find the option to provide services.

Choose your service categories from the list LinkedIn gives you. Pick the ones closest to what you genuinely sell, and be specific. A consultant should choose consulting and the exact areas they cover, not a scattergun of everything vaguely adjacent. Specific is what gets you found by the person who needs that one thing.

Write the description in plain language. You get room to say what you offer, so use the words a client would use to describe their problem, not the words you would use to describe your process. Nobody searches for your methodology. They search for their headache.

Set your location or mark yourself as remote, so LinkedIn and any human reader can see who you serve and where.

Then turn on the request button. This is the step people skip, and it is the whole point. The value sits in giving someone a direct route to reach you. If the button is off, you have built a shop window with the door locked.

That is the entire job. You now have a working enquiry channel that most of your competitors have left switched off.

Do not choose categories because you can do them. Choose them because you want to be hired for them.

Use this quick test: if someone sent an enquiry for this service tomorrow, would you be pleased to receive it, clear on what to sell them, and able to deliver it well? If the answer is no, leave it off.

Start with your core offer, the two or three adjacent services that naturally lead into it, and the language your best clients already use.

Example 1 LinkedIn Coach: I help consultants and professional service firms turn LinkedIn into a clearer route to visibility, credibility and qualified conversations. Support includes profile positioning, LinkedIn strategy, content systems and team training. When you enquire, tell me what you are trying to achieve, who you need to reach and when you want to begin your campaign.

Example 2 - Financial Planner: I help professionals, business owners and families make clearer, more confident financial decisions for the life they want to build. Support includes cash-flow and wealth planning, retirement planning, investment strategy, protection and long-term financial decision-making.

When you enquire, tell me what you want your money to enable, the financial decisions you are facing and when you would like to begin.

Example 3 - Business Coach

I help founders, consultants and service-business leaders build more focused, profitable and sustainable businesses. Support includes business strategy, offer development, pricing, sales and marketing systems, leadership support and accountability.

When you enquire, tell me what you want to improve, where the business feels stuck and when you want to begin.

OR

I help ambitious business owners move from reactive day-to-day work to clear priorities, stronger leadership and sustainable growth. Support includes strategic planning, decision-making, team leadership, commercial focus and practical accountability.

When you enquire, tell me the outcome you are aiming for, the main challenge you are facing and when you would like to get started.

➡️ Set a response rule before you publish

Do not switch this on and check it once a month. Decide now that you will review new requests every working day and reply within one business day where possible.

A simple first reply is enough:

“Thanks for getting in touch. I’d be happy to explore this. Before I suggest the right next step, could you tell me your goal, deadline, budget range and who will be involved in the decision?”

The feature gets the conversation started. Your follow-up determines whether it becomes work.

The mistake that wastes all of it

Setting it up, then forgetting it exists.

Treat reviews as evidence, not decoration

Ask for a review when the client has just experienced the outcome - not six months later, when the detail has faded.

Make the request easy to answer. Instead of “Would you leave me a review?”, ask:

“Would you be happy to leave a short LinkedIn service review? It would be especially helpful if you could mention the problem we started with, the work we did together and the result or change you noticed.”

They are other people saying they trusted you and were glad they did, which is the only proof a prospective client actually believes.

An empty Services section still helps, because the “Providing services” signal keeps working in the background regardless. A Services section carrying real reviews closes the deal while you sleep. The habit to build is small. Finish a job, ask for the review while the work is fresh, move on.

The part I want you to remember

LinkedIn keeps inventing features that ask for your money. This one asks for ten minutes and gives you a standing line for enquiries in return. It is sitting on your profile right now, switched off, doing nothing for you.

My friend turned hers on that afternoon. She had her first enquiry through it within the fortnight, from someone she had never met and would never have reached through posting alone.

Go and turn yours on this week. Then watch who walks through the door you just unlocked.

If reading The Link Tank has you thinking about your own LinkedIn, there are two ways forward. Do it yourself, using everything I share here or have it done for you.

Luminaries is the done-for-you route. My team and I build your thought leadership strategy, write and publish three posts a week in your voice, and grow your network with the right people. You stay focused on your work while the right opportunities start to find you.

No ad spend. No gimmicks. Everything grows from your own profile.

We are onboarding for September and places are limited because the work is hands-on.

See how it works →

Apply for Link Tank Luminaries

Then I saw how many people had shared it and how many were treating it as a plan rather than an opinion. So here we are.

Gary Vaynerchuk posted on LinkedIn this week telling people that most of them should be publishing far more often, potentially eight to 15 times a day. His reasoning was that the feed never shows every follower everything you publish, so nobody experiences it as spam. Everyone sees a fraction, the people who genuinely like your work see more and the algorithm sorts out the rest.

He has built a remarkable business, is a true entrepreneur, and I have learned plenty from watching him work. About LinkedIn, though, the advice is just wrong, and I want to explain why before anyone feels anxious and rebuilds their content calendar around it.

Nobody writes eight to 15 considered posts a day. What actually happens is that something gets automated and readers notice. You end up publishing more while saying less and the people you most want to reach are the quickest to spot it.

There is also the question of what happens after you press publish: The value on LinkedIn sits in the conversation underneath the post. That is where people work out whether they want to talk to you. Produce at that volume and you have no time left to reply to anyone properly, so you have traded the part that generates business for the part that only generates activity.

Your posts compete with each other, too. Publish again a few hours later and you shorten the life of the one before it. You are splitting your reach rather than adding to it. My advice would be to repost your post 7-8 hours later to resurface it in the feed instead.

For senior professionals, particularly in regulated sectors, there is a reputational cost on top of all that. Your profile is often the first thing a prospective client, a regulator, or a future hire looks at. Fifteen posts a day tells them something about your judgement that you probably did not intend.

So what is actually working? Two or three well-made posts a week, published consistently, with real time spent replying in your own comments and commenting thoughtfully on other people's. There is a reason that I schedule my posts when I am around for the following hour to respond to comments on them!

Analyse what works for you - I find Saturday morning posts have consistently gained the highest reach over the last 365 days, and all the posts which have “I” in the first line (are written in the first person) have gained 10k plus impressions.

That is the whole strategy. It is not exciting but it is what generates conversations that turn into work.

The wider point matters more than the specific advice. Being brilliant in one arena does not make someone right about another. Gary knows business. It seems that he does not know how LinkedIn behaves.

Here is what bothers me most about it: Gary will be fine either way. He has a media team, a production operation, and an audience large enough to absorb any strategy he chooses to test. The person who takes this advice at face value and spends the next month posting fifteen times a day is the one who finds out it does not work. By then they will have spent a month of their own credibility finding out, in front of exactly the people they were trying to impress.

That is the real cost of advice given confidently by someone who is not going to be around when it fails. Read the source, check the date and back your own judgement.

Warmest wishes as always,

Message me

It had come from LinkedIn. Not a website, not a directory. LinkedIn.

That sent me into the research on how AI search now works, and it ended with me rewriting my headline, my About section, and the way I publish, around one idea: making my profile and my writing genuinely quotable by a machine.

What follows is everything I learned and the exact audit I ran, shared in full, so you can do the same to be cited in searches.

LinkedIn is now the most-cited source in AI search for professional questions. When someone asks ChatGPT, Claude, Perplexity or Gemini who to trust in your field, the answer is increasingly pulled from LinkedIn. That single fact should change how you think about every profile, post and article you publish this year.

Here is the 30-second version if you are reading this on your phone between meetings:

• LinkedIn is now the top-cited domain in AI search for professional queries

• LinkedIn articles, the long-form publishing feature, take 72.2% of all AI citations of LinkedIn content, and they are the format most people ignore

• Likes, comments, hashtags and emojis show near-zero correlation with getting cited by AI

• The format that gets cited is a 500 to 2,000 word article, written the way a real expert would speak, with a clear claim in the first two sentences.

• If your visibility strategy still ends at “post three times a week”, you are optimising for the last decade.

  Subscribe now

The conversation nobody is having

For 15 years, the LinkedIn conversation has centred on one thing. The algorithm. How to crack it. How to write hooks that stop the scroll. How to squeeze more reach out of the feed before it eats you.

That conversation is not wrong, it is just no longer the most important one. While everyone has been arguing about dwell time and carousel formats, the way people find experts has changed underneath our feet.

When your ideal client wants to know who to hire, who to trust, or who understands their problem, a growing number of them are no longer typing into Google. They are typing into ChatGPT, Perplexity, Claude, or the AI panel that now sits at the top of Google itself. The answer that comes back is, more often than not, sourced from LinkedIn.

What “good content” now means

Here is the part that will feel uncomfortable, especially if you have spent years learning to write short, punchy, hook-heavy posts. The content AI cites is not the content that goes viral.

A recent study by OtterlyAI found four things worth your attention.

1. One. LinkedIn articles take 72.2% of all cited LinkedIn content. Regular feed posts, the format everyone obsesses over, account for around 26%. The format most professionals abandoned years ago is now the one doing the heavy work.

2. Two. Vanity metrics do not translate. Likes, comments, hashtags and emojis all showed near-zero correlation with AI citations. The numbers you have been chasing on the feed have no measurable connection to whether a machine quotes you.

3. Three. Video reduces citation likelihood. Posts with video averaged 5.85 citations against 7.87 for text-only posts. AI reads text. It cannot quote your face.

4. Four. Length matters. The median cited article ran to 1,021 words. The pieces getting cited cluster in the 500 to 2,000 word range, not the 150-word feed post.

None of this is what the top LinkedIn creators have been teaching. The viral formula rewards short, sharp, emotive text. The AI-citation formula rewards clear, structured, expert long-form. These are two different games, and from here you probably need to play both.

What this means for your visibility

Three consequences. I want you to think each one through before you decide what to change.

Your LinkedIn profile is now an AI training input, not a CV. Every claim in your headline, every keyword, every case study in your Featured section is being read, indexed and potentially quoted back to your future clients by a machine. Vague, buzzword-heavy profiles will not be cited. Specific ones will.

Long-form is back, and it looks nothing like the long-form of 2018. The LinkedIn article you have been ignoring for years is now the most powerful single unit of content on the platform. Not because LinkedIn is pushing it, but because AI is quoting it.

If you already run a LinkedIn newsletter, this is you. A newsletter is built on the article engine. Every edition publishes as an article at its own public web address on the /pulse/ path, indexed and readable whether or not someone subscribes . The subscriber email and notification are added on top. Your newsletter is already the citable format, and it reaches your subscribers in the same move. The one condition is that it stays public rather than gated, because a piece a machine cannot open is a piece it cannot quote.

The people who lead the next five years on LinkedIn will not be the loudest but they undoubtedly will be the most citable.

It is closer to writing a good briefing note than a good social media post.

How to make your LinkedIn content AI-citable this week

An audit you can run in an afternoon. It applies to standalone articles and to newsletter editions in exactly the same way, because they are the same format underneath.

1. Rewrite your headline as one citable sentence. Not a string of buzzwords separated by pipes. One clear sentence naming who you help, what you do, and one specific proof point. The headline is the line an AI is most likely to quote back word for word when a client asks who the best expert on your subject is. Write a sentence you would be happy to see repeated.

2. Fill your About section with claims a machine can verify. Replace “passionate about” language with named numbers, named publications, named credentials, and named clients where you have permission. A machine cites what it can check. If a claim cannot be verified, treat it as invisible.

3. Publish one long-form piece a month, minimum, and count your newsletter. Between 500 and 2,000 words. A newsletter edition qualifies, so you may already be doing this. Publish on LinkedIn first, before your own blog, so the citable copy sits on the domain AI now trusts for professional queries. Use clear subheadings so a model can extract each section cleanly.

4. Lead with your claim, then define your terms. Put the central point in the first two sentences, where both readers and models look first. Then define the thing you are known for, in plain form. “X is Y because Z.” Machines quote clean definitions. Vague expertise gives them nothing to lift.

5. Add a short question-and-answer section. This is the step most professionals miss. AI models pull question-and-answer pairs close to word for word, because the format matches how people ask questions in the first place. Anticipate the exact questions a client would type, such as “how much does X cost”, “how long does Y take”, or “what is the difference between A and B”, and answer each in two or three clear sentences at the foot of the piece. This is a principle from Answer Engine Optimisation rather than a figure from the OtterlyAI study, and it is well established across AI-search guidance. A short set of frequently asked questions turns one article into several separately quotable answers.

6. Make your profile and your content say the same thing. Models cross-reference. If your headline claims one specialism and your articles cover another, the machine reads mixed signals and trusts neither. Consistency across your profile, your articles and your newsletter is what builds an identity a machine can repeat with confidence.

7. Stop optimising for the algorithm alone. Keep writing feed posts. Keep engaging. Recognise the division of labour: the feed post gets you seen this week, and the long-form piece gets you cited for the next three years.

8. Get quoted somewhere other than LinkedIn. A machine trusts a name it meets in more than one place. A guest article, a podcast, a mention in someone else’s piece, an industry roundup- each one is another independent source agreeing that you are the expert. One strong profile plus a scatter of outside mentions is far more citable than a profile standing on its own.

9. Keep your best pieces current. Recency counts. AI-search results lean heavily towards recently published or recently updated pages, so a strong article from two years ago is worth revisiting. Refresh the figures, add the current year to the title where it fits, and republish. A current date on a solid piece does a surprising amount of work.

10. Make your profile and your content say the same thing. Models cross-reference. If your headline claims one specialism and your articles cover another, the machine reads mixed signals and trusts neither. Consistency across your profile, your articles and your newsletter is what builds an identity a machine can repeat with confidence.

11. Stop optimising for the algorithm alone. Keep writing feed posts. Keep engaging. Recognise the division of labour: the feed post gets you seen this week, and the long-form piece gets you cited for the next three years.

Where the SEO settings actually live

Both Substack and LinkedIn let you set a separate Search Engine Optimisation title and description, which is what Google and the AI engines read instead of your display title. Most writers never touch either as they are pretty well hidden 🫤.

Here is where to find them:

On a LinkedIn article or newsletter edition:

1. Open your draft in the article editor.

2. Click the pulldown at the top of the editor and select the Manage menu, then click Settings.

3. In the SEO title field, write a keyword-led title of up to 60 characters. Search engines will show this in place of your display headline, so put the phrase people actually search for at the front.

4. In the SEO description field, write 140 to 160 characters summarising the piece, with your main keyword included naturally. LinkedIn caps this field at 160 characters.

5. Publish as normal. Your subscribers still see your display headline; only search engines see the SEO version.

   

On a Substack post:

1. In your draft, open the post settings from the button at the right bottom of the editor, then find the SEO settings.

2. Set the SEO title, keeping it under 60 characters so it does not get cut off in results.

3. Set the SEO description. If you leave it blank, Substack pulls your subtitle in automatically, so a strong subtitle is your safety net, but a purpose-written description almost always does the job better.

4. Check the post URL while you are there. Keep it short and keyword-led. Change it before publishing, never after: editing the URL after publication breaks every existing link to the piece (Substack, SEO guide).

5. Set the social preview image in the same panel, then publish.

The whole exercise takes five minutes per piece. It is the difference between your display title, written to be clicked by a human, and your search title, written to be found by a machine. You are allowed both.

Questions you might be asking

1. Does my LinkedIn newsletter count, or do I need to write separate articles? It counts. A newsletter edition publishes as a LinkedIn article at its own public web address, so it is indexed and citable in exactly the same way. Keep it public rather than gated and you are already doing the work.

2. Should I stop writing feed posts? No. Feed posts still get you seen and keep relationships warm. The change is recognising they do a different job: the feed post reaches people this week, and the article gets quoted for years.

3. How long should an article be to get cited? Between 500 and 2,000 words. The median article cited by AI models ran to 1,021 words (OtterlyAI, 2026), so aim for that neighbourhood rather than either extreme.

4. Do likes and comments help me get cited by AI? The evidence says no. Likes, comments, hashtags and emojis all showed near-zero correlation with AI citations. Engagement is worth having for other reasons, but it is not what earns the quote.

Everything above was checked against the sources when I wrote yesterday (9th August 2026). AI search is moving quickly, so if you are reading this much later, verify the numbers before you rely on them.

The uncomfortable question

If a decision-maker in your ideal client’s business types your name, your niche, or your exact service into ChatGPT tomorrow morning, what comes back?

For most professionals I speak to, the honest answer is not much, or nothing, or worse, a competitor. That is the visibility gap of 2026. It is not about follower count. It is not about how often you post. It is about whether the machines now rewriting how people find experts have anything specific to say about you.

Almost no one is fixing this yet. The window is wide open, and it stays open for exactly as long as it takes the rest of your industry to work it out. Judging by how fast this is moving, that is not very long.

If this was useful, subscribe to get The Link Tank each week. It is free, and always focused on the how.

Subscribe now

You open LinkedIn to grow your network. You send a few connection requests. You accept a few more. The number next to your name creeps up and it feels like progress.

Except your accountant will tell you otherwise.

You are connected to hundreds of people you have never once spoken to. No message. No conversation. No client relationship. No revenue.

So what have you gained aside from a public number?

Let’s step back.

Those people have a name. They are your dormant connections, and for most senior professionals, they are the largest untapped opportunity on the platform.

This article shows you how to find every one of them, and how to reopen the conversation in a way that feels human. There is a name for the method and a set of templates you can lift straight out at the end.

Why your dormant connections outperform your next 500 new ones

The usual instinct, when LinkedIn feels flat, is to add more people. More connections, more followers, a bigger number. It rarely works, because a brand new connection knows nothing about you. You have to earn the right to a conversation from a standing start.

A dormant connection is different. At some point, one of you accepted the other. There was a reason. A shared event, a mutual contact, a company they were interested in, a piece of your content that caught their eye. That reason is a foothold, and it gives you something true to reference. A message with a true reference behind it gets a reply. A generic one does not.

There is a trust factor on top of that. When you appear in someone’s inbox as an existing first-degree connection, you are not a stranger arriving uninvited. You are someone they already chose to connect with. That single fact changes how the message is received before you have written a word.

The Dormant Connection Reset

Here is the framework the rest of this article hangs on. I call it the Dormant Connection Reset, and it has four moves: surface, sort, reopen, nurture.

• Surface means pulling your own LinkedIn data and finding everyone you are connected to but have never messaged.

• Sort means grouping that list so a past client and a near-stranger get different treatment.

• Reopen means sending a short, warm, no-pitch message that restarts the relationship.

• Nurture means following through when they reply, without lunging at the first opening.

You can run the whole system in an afternoon, then repeat it every few months as your network grows.

How to run it this week

Step one: Get the data

LinkedIn lets you download a copy of your own data, including your full connections list and your entire messaging history. You need both.

LinkedIn changes its menus from time to time, so look for the nearest equivalent if a label has moved.

1. Click your profile photo, the Me icon, at the top of LinkedIn.

2. Choose Settings and Privacy.

3. In the left-hand menu, select Data privacy.

4. Click Get a copy of your data.

5. You will see two options.

6. Choose Download larger data archive and request that too. It arrives as messages.csv, usually within 24 hours, occasionally longer for a large or old account.

   

These files contain other people’s personal information. Handle them with the same care as any client data. If you use an AI tool for the cross-reference, use a paid business or enterprise plan with a data processing agreement in place, not a free consumer account, and check your firm’s policy first. Delete the files and the chat when you are finished.

Step two: find the people you have never messaged

Once both files have arrived, open Claude and upload Connections.csv and messages.csv together. Then give it a clear instruction. Here is one you can copy and adjust:

I’ve uploaded two files from my LinkedIn data export: Connections.csv, my full first-degree connections list, and messages.csv, my complete LinkedIn messaging history. Cross-reference them. Give me a list of every connection in Connections.csv who does not appear anywhere in messages.csv, either as a sender or a recipient. These are my dormant connections, the ones I have never messaged. Present them as a table with name, company, position and the date we connected, sorted by most recent first. Where a profile web address is available in both files, match on that rather than the name to avoid errors, and flag anything you were unsure about.

I’ve uploaded two files from my LinkedIn data export: Connections.csv, my full first-degree connections list, and messages.csv, my complete LinkedIn messaging history. Cross-reference them. Give me a list of every connection in Connections.csv who does not appear anywhere in messages.csv, either as a sender or a recipient. These are my dormant connections, the ones I have never messaged. Present them as a table with name, company, position and the date we connected, sorted by most recent first. Where a profile web address is available in both files, match on that rather than the name to avoid errors, and flag anything you were unsure about.

That last line in the above prompt matters. Names are not unique, and people change them, so matching on the profile web address is safer where the export includes it. Ask for uncertain matches to be flagged rather than guessed.

➡️What comes back is your dormant list. For most established professionals, it runs into the hundreds. That number tends to surprise people, and it is the whole point. Every row is a conversation you already have permission to start.

Share

This newsletter is free because people like you share it with their network who also need this

Step three: Sort before you write

Do not treat the list as one block. A past client and a conference contact from three years ago need different opening lines.

Be realistic about what the data can do here. Your export holds each person’s name, company, position and the date you connected. From that, Claude can group people by industry, by seniority, and by how well they match your target audience. It can also spot clusters of connections made around the same date, which often point to a single event or a post that did well.

What the data cannot tell Claude is your history with someone. It has no way of knowing that a name was a client three years ago, or that you met at a particular dinner, because none of that is in the file.

So you supply that part: Paste in a short list of your past clients or their companies, and the names of events you have spoken at or attended, and ask Claude to tag anyone on your dormant list who matches. Between the data it can read and the context only you hold, you end up with four workable groups: past clients and former colleagues, event or mutual-contact connections, people in your target audience who connected after your content, and cold adds where neither of you can remember why.

💡 There is a way to make the sort sharper still, if you use Outlook or another email account for work. Your LinkedIn export cannot see your history with someone BUT your inbox can!

Connect Claude to your email, or give it an export of your contacts and sent items, and it can cross-check your dormant LinkedIn list against the people you have actually corresponded with. Someone you have exchanged twenty emails with is not a cold connection, whatever LinkedIn says. That match shows you where a real relationship already exists, and it places people in the right group with far more confidence than company and job title alone.

Here is a prompt you can use once your email is connected.

Here is my dormant LinkedIn connections list, the people I am connected to but have never messaged on LinkedIn [paste the list or attach it]. Using my connected email account, check each person against my email history, matching on name and email address where you can. For anyone I have corresponded with, tell me roughly how many emails we have exchanged and when we last spoke, and mark them as a warm relationship rather than a cold connection. Then re-sort the whole list into four groups: past clients or people I clearly have a working relationship with, event or mutual-contact connections, target-audience connections, and cold adds where neither of us can remember why we connected. Flag anyone you were unsure about rather than guessing.

Here is my dormant LinkedIn connections list, the people I am connected to but have never messaged on LinkedIn [paste the list or attach it]. Using my connected email account, check each person against my email history, matching on name and email address where you can. For anyone I have corresponded with, tell me roughly how many emails we have exchanged and when we last spoke, and mark them as a warm relationship rather than a cold connection. Then re-sort the whole list into four groups: past clients or people I clearly have a working relationship with, event or mutual-contact connections, target-audience connections, and cold adds where neither of us can remember why we connected. Flag anyone you were unsure about rather than guessing.

One word of caution here, particularly if you work in a regulated field. Connecting an inbox gives the tool access to correspondence that often contains confidential client material. That is a heavier data decision than exporting a LinkedIn file, and some firms will not allow it. Check your own rules before you connect anything.

Start with the first two groups. The shared reference does the work for you. The last group deserves a lighter touch, and some of them are fine to leave dormant ;)

Step four: reopen with a real hello

Now you write, one message at a time, following a few principles that keep this human rather than salesy.

Lead with the connection point, not yourself. The first line should reference the real reason you are linked, which proves the message was written for them. Give before you ask, and often do not ask at all, because the strongest re-openers carry no request. Make it effortless to reply, since one easy question beats a paragraph of context. Keep it to two or three sentences, written the way you speak. Be honest about the gap, because a light acknowledgement that it has been a while is disarming and true.

The templates for each group are at the end, ready to lift.

Subscribe now

Step four: reopen with a real hello

Now you write, one message at a time, following a few principles that keep this human rather than salesy.

Lead with the connection point, not yourself. The first line should reference the real reason you are linked, which proves the message was written for them. Give before you ask, and often do not ask at all, because the strongest re-openers carry no request. Make it effortless to reply, since one easy question beats a paragraph of context. Keep it to two or three sentences, written the way you speak. Be honest about the gap, because a light acknowledgement that it has been a while is disarming and true.

The templates for each group are at the end, ready to lift.

A worked example

Picture a commercial solicitor who came through one of my training programmes.

She ran her export and found more than 1,400 dormant connections. Fourteen hundred people she was connected to and had never once messaged. The number alone changed how she thought about her network. When she sorted the list, one group stood out: around 200 people she had met at conferences and industry events and never followed up with.

She started there. A two-line message to each that referenced the event and asked one real question about their work. She sent them in small batches over a month, written by hand, spaced across her days.

Close to half replied. Most were simple exchanges at first, a bit of catching up, a comment on a shared contact. Around a dozen of those conversations, over the following weeks, opened into something with real substance. Three led to live pieces of work, and one became the largest instruction her team took on that quarter. She did not sell to anyone. She reopened relationships, and the work came from the conversations, not from any ask.

Where people go wrong

Three failure modes account for most disappointing results.

1. The first is speed. People find 600 names and try to clear them in a day with copy and paste. The message stops being personal, the replies dry up, and the exercise teaches the network that you send noise. Batches of ten or fifteen, written properly, beat a hundred fired off at once.

2. The second is the early pitch. Someone replies warmly, and the temptation is to pivot straight to your services. That is precisely the move that trains people to stop replying. The reopening is not the sale. It is the return of a conversation, and the reason to talk further has to surface on its own.

3. The third is skipping the sort. A message that would land well with a former colleague reads as odd when sent to a near-stranger. The four groups exist so the words match the relationship. Spend the ten minutes to sort, and every message afterwards is easier to write.

The templates you can lift

Square brackets are prompts for you, not words to send. Send these one at a time, written and checked by you.

The core re-opener, with no pitch:

Hi [First name], we connected back in [month or year, or after the [event] event] and I don’t think we’ve actually spoken since. I’ve been [meaning to reach out / following your work at [company]] and wanted to say a proper hello. How are things going with [something specific and true to them]?

For a past client or former colleague:

Hi [First name], it’s been far too long. I still think about [specific project or moment you shared]. I’d love to hear what you’re working on these days. Are you still at [company], or has something new come along?

For an event or mutual-contact connection:

Hi [First name], we connected around [event or through [mutual contact]] and I’ve been meaning to reconnect properly ever since. I really enjoyed [something specific from that context]. What are you focused on at the moment?

For someone who connected after your content:

Hi [First name], thank you for connecting a while back. I noticed we’re both [shared interest or field] and I realised we’ve never actually spoken. What’s the biggest thing on your plate this quarter?

The speedy one for quicker results
Hi [First name], we connected a while back but never got the chance to talk properly, which I know is on me but feels like a shame. Would you be up for twenty minutes in the diary in the last week of [this month], if you are around? [optional add:I'd like to get to know you and what you're building, with no agenda beyond that.]
Warmest, Your Name

The highest converting one for my clients

Hi [First name], I hope you’re well and you’re enjoying the Summer. As a fellow X, could we please have a 1-to-1 to explore how we may be able to help each other in business? Regardless of my work, I like to cross-refer and collaborate where possible and indeed I encourage all my clients to do the same.

Warmest, Your Name

The full method, steps and templates, is in a one-page checklist here if it's useful:

Download here

The bottom line

Your network is the set of conversations you are actually having, whatever the number on your profile says. Most professionals have hundreds of connections and a handful of conversations, then try to fix the gap by adding more connections. The faster route runs the other way.

Surface the people you have never messaged. Sort them. Say a real hello, one at a time, and follow through when they reply. The opportunity was already sitting in your network. You simply had not spoken to it yet.

If you’d rather not run this alone

Reactivating dormant connections is one part of the managed LinkedIn business development programme I run for time-poor individuals and firms.

The first step is a paid strategy call, where I provide a written profile review in advance and then we look at your situation and work out together whether the programme is a fit. There is no obligation beyond the call, and we cover exactly what the programme involves and what it costs while we are on it.

If that is useful, you can book the call here.

Book a Call

One last question before you go. When you run your export, which group surprises you most, the old clients or the strangers you cannot place?

TLDR: The EU AI Act’s transparency rules under Article 50 took effect yesterday, 2 August 2026, and they cover AI-assisted content seen in the EU regardless of where you posted it from.

China, India, South Korea and several US regimes already impose their own disclosure duties. LinkedIn tags some AI images and demotes generic AI posts, but the legal obligation to disclose sits with you, not the platform.

This issue explains what Article 50 requires, who it reaches, and gives you a five-step compliance plan plus disclosure wording you can use today.

Since yesterday, your LinkedIn posts fall under a law LinkedIn has not built for.

On 2 August 2026, the EU AI Act’s transparency rules finally switched on. They apply to AI-generated content seen in the EU, even if you posted it from London, New York or Singapore. At the same time, a wider wave of AI-content disclosure laws in the US, Asia and beyond means your LinkedIn presence is now judged under multiple regimes at once.

LinkedIn has started tagging some AI-generated images and demoting generic AI posts. It has not built a complete legal compliance layer for you. Demotion and platform labelling are platform policies. Disclosure is a legal obligation, and that obligation now sits with the people and firms who create and publish AI-assisted content, not with LinkedIn’s feed algorithm.

Why 2 August matters worldwide

You have probably seen “2 August 2026” in AI headlines, often framed as the day the AI Act starts. Much of that commentary is slightly off. The EU’s Digital Omnibus package, which reached provisional political agreement on 7 May 2026 and was signed into law on 8 July 2026, pushed the high-risk system duties out to December 2027 and August 2028. One article was left out of that deferral: Article 50, the transparency obligation.

Article 50 applies from 2 August 2026 and does three things that matter for LinkedIn and other social platforms.

• It requires providers of AI systems designed to interact directly with people to tell users when they are dealing with AI. That covers chatbots, assistants and social bots.

• It requires generative AI providers to mark synthetic content, including images, video, audio and text, in a machine-readable way so downstream platforms and tools can detect that it is AI-generated.

• It requires deployers, meaning users of AI systems, to disclose certain uses to individuals: emotion recognition, biometric categorisation, deepfakes and AI-generated text that informs the public on matters of public interest.

There is a short grace period until 2 December 2026, but it is narrower than most coverage suggests.

It applies only to the machine-readable marking of outputs from generative systems already on the EU market before August. Chatbot disclosure, deepfake labelling and public-interest text transparency have applied in full since yesterday. One further detail from the European Commission’s final guidelines, adopted on 20 July 2026, is worth knowing: content created before 2 August does not need labelling, but for AI-generated public-interest text the relevant date is publication. Text generated in July and published in August must be disclosed.

LinkedIn is a global platform. If any segment of your audience sits in the EU, Article 50 can apply to your AI-assisted posts there, regardless of where you are based.

The wider wave: disclosure laws beyond Brussels

Treating AI disclosure as an EU matter misses the bigger picture. By mid-2026, a growing list of jurisdictions have passed or are enforcing AI-content disclosure laws that catch brands, influencers and professionals with worldwide audiences.

In the United States, the Federal Trade Commission (FTC) expects clear disclosure of both sponsorship and material connections, which increasingly includes AI involvement in endorsements and advertising campaigns. Several states, including New York and Colorado, have specific rules for synthetic performers, political ads and high-risk AI deployers.

China has required labelling of AI-generated content since 1 September 2025 under the Labelling Measures, which build on the deep synthesis regulations in force since 2023. Both visible labels and embedded metadata are mandatory, with strict platform enforcement and obligations placed on users themselves to declare synthetic content when they publish it.

India amended its Information Technology (IT) Rules in February 2026 to require labelling of AI-generated or modified content, and the framework gives authorities takedown powers.

South Korea’s AI Basic Act came into force on 22 January 2026 and requires labelling of AI-generated content, with fines deferred for roughly the first year of the regime. Other APAC markets are beginning to treat unlabelled synthetic content as deceptive under advertising law.

On top of that, major platforms including TikTok, Meta and YouTube have rolled out “Made with AI” labelling systems that apply globally, regardless of where you post from. LinkedIn has moved more slowly, focusing on demotion and partial image tags rather than a full cross-format AI label.

If your influence and client base are worldwide, the practical rule is simple. Assume you are inside at least one AI-disclosure regime at all times, and design your practice accordingly.

Open The Global AI Disclosure Tracker

What Article 50 actually asks of you

The European Commission’s July 2026 guidelines and the official Code of Practice on transparency, assessed as adequate by the Commission and the AI Board on 8 and 9 July, spell out Article 50 in practical terms. Four use-cases matter for professional content.

1. Direct interaction: chatbots and assistants

Any AI system that interacts directly with individuals must clearly inform them they are dealing with AI. This applies even when the chatbot is embedded in a website, app or customer-support channel and the brand is human-facing.

Drafting assistance sits on the other side of the line. Using AI to draft a message that you review and send yourself, including LinkedIn’s own built-in AI features, does not create an AI system interacting directly with anyone. The human presses send, so the chatbot disclosure duty does not arise. The dividing line is autonomy. The moment a system responds to people without you in the loop, disclosure duties attach, and on LinkedIn itself the User Agreement stops you before the AI Act does.

2. AI-generated content: images, video, audio and text

Article 50 requires providers of generative AI systems to add machine-readable marks to synthetic content so downstream detection is possible. Deepfakes and realistic synthetic media of real people must also be clearly and visibly labelled.

For you as a deployer and publisher, the practical effect is this. If you post AI-generated or heavily AI-edited images, especially realistic portraits or scenes, you should assume they will carry provenance metadata under the Coalition for Content Provenance and Authenticity (C2PA) or International Press Telecommunications Council (IPTC) standards, and you should add a visible disclosure in captions. If you share AI-generated video or audio that depicts or imitates real people, clear disclosure that it is synthetic is expected.

For AI-generated text, the Commission’s guidelines focus on public-interest use. Text published to inform the public on matters of public interest, such as elections, health or systemic risk, without human editorial control must be labelled as AI-generated. Routine, low-stakes marketing copy is less likely to trigger the text labelling requirement. The more your AI-authored posts look like public-interest commentary, the closer they move to that line.

3. Emotion recognition and biometric categorisation

Deployers must inform individuals when AI systems process their biometric data for emotion recognition or categorisation, whether in real time or retrospectively. This is less common on LinkedIn. It becomes relevant if your firm uses AI tools to analyse facial expressions or voice tone in video calls, webinars or recorded events, and then acts on those insights.

4. Deepfakes and public-interest text

Deepfakes, meaning synthetic audio-visual content depicting real people, must be clearly labelled as such. Failing to do so is a breach of Article 50. Likewise, AI-generated text on public-interest matters published without human editorial control must be visibly disclosed.

For senior professionals whose thought leadership touches on policy, regulation or macro-finance, this means two things. If you use AI to draft or fully generate LinkedIn articles or posts about regulatory changes, elections, public health or systemic risks, and they are published largely as-is, you should tell readers that AI generated the text. If you ever experiment with synthetic video impersonating real executives, partners or public figures, that content requires clear deepfake labelling and is high-risk from a compliance perspective.

Why you cannot rely on LinkedIn to do this for you

LinkedIn has taken steps to respond to AI content.

Feed demotion comes first. Generic, low-signal AI content is increasingly suppressed, reducing its reach. That is quality and misinformation control, not the legal disclosure mechanism it is designed to be.

Image tags come second. LinkedIn applies automatic C2PA-based tags on some images where it detects AI involvement, based on provenance metadata and other signals. This is helpful but limited to images. It does not cover AI-generated video, audio or text.

There is no uniform AI label for text. Unlike some platforms, LinkedIn does not currently auto-label AI-generated posts or articles.

Article 50’s duties fall on providers and deployers, the organisations and professionals who use AI systems and publish their outputs, rather than on LinkedIn as a neutral distributor. LinkedIn can help but it is not your legal shield.

Territory versus audience: how the law reaches you

Because your audience is worldwide, you need to think in audience geography, not company registration. The EU AI Act follows the General Data Protection Regulation (GDPR) model: it can apply when AI output reaches individuals in the EU.

This is where the AI Act parts company with the analogy most people reach for. GDPR requires targeting and mere accessibility of a website has not been enough to trigger it. Article 2(1)(c) of the AI Act contains no targeting requirement. It asks only whether the output is used in the Union.

On LinkedIn, EU-based followers reading and engaging with your post is use of the output in the Union on any ordinary reading. For anyone with EU connections, followers or commenters, and that means most professional networks of any size, I would treat scope as engaged in fact rather than in theory. The realistic exception is a network genuinely confined to one non-EU country.

I should be transparent about the limits of my reading. The “output used in the Union” test has not been judicially interpreted. The Commission’s guidelines address what Article 50 requires rather than how territorial scope maps onto social platforms. Reasonable lawyers differ on how far passive distribution counts. This is my reading of the provision, and it is not a settled one.

(US FTC rules and state-level laws apply where your content is aimed at US consumers, particularly in advertising and endorsements. China's labelling rules bite when synthetic content is distributed through Chinese platforms or reaches users there.)

The risk tier: this is a board-level issue

enforced by national authorities, who gained those powers on 2 August. FTC penalties for deceptive practices and failure to disclose material facts in advertising are similarly serious.

Two features move this issue from “social media” to the board agenda:

1. The obligations attach to providers and deployers, that is, to organisations and professionals who design, adopt and use AI systems in customer-facing contexts. Marketing, business development and client communications fall directly into that scope.

2. The same AI-assisted assets you post on LinkedIn are often repurposed on websites, newsletters, investor decks and webinars, multiplying exposure across channels and jurisdictions.

Ignoring AI disclosure because “it’s only LinkedIn” is therefore risky. An unlabelled AI deepfake or public-interest article on LinkedIn can be the entry point into a broader investigation of your AI content practices.

Five practical steps to take this week

You cannot redesign LinkedIn’s product overnight. You can close the gap on your side. Here is a practical, globally relevant plan:

Read more

Here is my process, in the open, and why disclosure beats suppression.

Subscribe

I am Melanie Goodman, a lawyer turned LinkedIn consultant and the writer of The Link Tank. I teach time-poor professionals to use LinkedIn and Substack intelligently, without hacks or hype. If a platform announcement has changed how you feel about your own writing this month, you are in the right place.

Before I ran a consultancy, I practised law.

My first weeks as a junior associate at Allen & Overy taught me a lesson I have never shaken off. I sent a partner my first attempt at a drafting exercise and it came back with more red ink than black. Nobody apologised for that. Nobody expected a first draft to be the finished article. The mark-up was the method.

The habit that stayed with me was the blackline. Every revised document went out with a comparison version showing exactly what had changed since the last draft. The other side did not have to trust us. They could see the edits for themselves, line by line.

I spent sixteen years working across borders after that, and the rule held everywhere. You did not hide the changes. You showed your working.

Yes, this essay is about writing with AI.

It is also about the strangest fact of my working life: a lawyer, trained to trust nothing unverified, now writes with a tool famous for inventing things.

I use Artificial Intelligence in my research and my drafting. I have said so before and I will keep saying so. What I want to examine this week is what the two platforms I work on have decided to do about it. They have chosen opposite designs, and the difference matters more than either announcement admits.

How I actually use AI

My process has a fixed order. I decide the angle. I am interviewed about scope before a word is drafted, because structural recommendations made before questions are usually wrong. A draft follows. I reject the parts that do not sound like me or do not say what I think. Statistics are checked against a named source before they go in, never after. Claims that cannot be verified are cut or bracketed for me to resolve.

The tool does the volume; I make the judgement. On a good week the draft survives two rounds of my corrections. On a bad week it survives five.

An example from this month: While drafting my anniversary issue, a scene appeared in the draft that I did not recognise, a conversation with a wealth adviser in Geneva that had never happened in the form described. The machine had written something plausible. Plausible was the problem. The scene was bracketed, checked and rewritten before a reader ever saw it and the rule it produced now applies to everything I publish: anything invented gets flagged for me to verify or replace. My old profession would call that a mark-up. I call it the job.

What Substack decided to show

I read the Substack announcement on my phone last Tuesday evening, somewhere between finishing client work and being asked what was for dinner. My first feeling was recognition rather than worry. A platform had reached for the thing my old profession has always relied on: showing the reader what changed. Then I kept reading, thought about LinkedIn, and realised the two platforms I work on had just answered the same question in opposite ways.

On 21 July, Substack switched on AI detection through an integration with Pangram. Readers on the web and the iOS app can scan any post, note, reply or comment over 100 words and see an estimate of how much was human-written, AI-assisted or AI-generated (Axios, July 2026). Writers can scan their own drafts before publishing, report scans they believe are wrong, and add a statement called How I Make This that readers see alongside the work. Substack’s chief executive, , framed the goal as stopping Substack becoming “like LinkedIn”.

LinkedIn faced the same problem two months earlier and chose a different answer, which is presumably why Best reached for the comparison.

Give Substack this much: the design is honest about its own uncertainty. The estimate sits in the open. The writer gets a right of reply beside it and a correction route when the scan is wrong. Readers decide what it all means.

What LinkedIn decided to do instead

In May 2026, LinkedIn announced new detection systems, trained with its editorial team, aimed at posts that appear machine-generated and carry no clear point of view (LinkedIn Editorial blog, May 2026, reported by The Decoder). Laura Lorenzetti, Vice President and Executive Editor of LinkedIn Global Editorial, set the standard plainly: “Your posts and comments need to represent your voice and your perspectives.”

The consequence is different from Substack’s.

Flagged content is not labelled but it is throttled. Posts judged generic lose distribution and stay largely within the author’s own network. LinkedIn reported a 94 per cent accuracy rate on generic content in early tests. That figure is self-reported, and the false positive rate has not been published, so treat it as a claim rather than a finding.

Images are treated differently: Since May 2024, LinkedIn has read Content Credentials, the metadata standard from the Coalition for Content Provenance and Authenticity (C2PA), and displays a small CR badge on images that carry it.

Clicking the badge opens a panel naming the tool that made the image, the date, and whether AI was involved. Two limits matter:

1. The system reads metadata rather than detecting anything, so a file with the metadata stripped out arrives unlabelled.

2. It also fires on a real photograph that has been through a single AI edit, which means a retouched headshot can carry the same badge as a fully generated scene.

The scale of the problem is significant.

A Pangram Labs study of more than a million social media posts, reported in July 2026, found that over 41 per cent of long-form LinkedIn content was fully machine-generated, the highest share of any platform studied.

The irony is also real. Microsoft, LinkedIn’s parent company, spent the same period promoting Copilot features that draft LinkedIn posts for you.

How to manage this on LinkedIn today

Two designs, one wrong measurement

Here is the difference that is of consequence

Here is the difference that matters.

Substack tells the reader and lets the reader decide.

LinkedIn splits the answer by format.

Your image gets a public label. Your words get a private penalty. Disclosure for pictures, suppression for prose, and no explanation for why the reader deserves the truth about one and not the other.

There is a further inversion worth noticing:

Substack lets a writer disable detection on their own work, and readers can draw their own conclusions from that choice.

LinkedIn’s image badge has no off switch at all. The platform that grades you trusts you with a setting. The platform that throttles you does not.

As a lawyer, I notice that this distinction is about to stop being philosophical.

Article 50 of the European Union’s AI Act is scheduled to take effect in August 2026. It requires disclosure to EU users when content has been machine-generated, with penalties of up to fifteen million euros or three per cent of worldwide annual turnover. Reducing the reach of a post is a moderation choice. Telling the reader what they are reading is a transparency duty.

The first does not obviously satisfy the second. A proposal to delay parts of the Act was tabled in late 2025, so watch the implementation date rather than assume it.

It cannot read the rejected drafts or the sentence that was cut because it could not be verified. A percentage describes the surface of a text. Whether a person stood behind the words is a different question and no detector answers it.

Interfaces create norms through what they measure and what they display. Substack’s norm will be disclosure with context. LinkedIn’s norm will be caution, because when the penalty for detection is invisible reach, writers will not experiment with the tools in the open. They will hide the use or stop the writing but neither outcome makes the feed more honest.

The image labels teach a different lesson. Because the badge reads metadata rather than the picture itself, the label is optional for anyone willing to strip a file before uploading. A rule that only binds the honest is a strange foundation for trust.

How I make this

I research with AI and I draft the first draft with it and this essay is its own example. AI gathered the platform coverage and produced the drafts. I set the argument, directed the corrections through several rounds and rewrote until the piece said what I think in the way I say it. When Substack’s scanner reads it, it will find the machine’s fingerprints on the sentences and mine on the personal experiences. Both readings are accurate. Consider this the blackline for the document you have just read.

Law worked out the answer to this decades ago. The blackline did not ask anyone to trust the drafter; it showed the changes and let the reader judge them. A percentage beside my writing does not trouble me for the same reason.

This essay was made the way I describe above. Substack can grade it. LinkedIn can decide what happens to the post that links to it. You can decide whether it earned your attention. Of those three judgements, only the last one was ever the point.

Would you keep using AI in your writing if every reader could see a percentage beside your name? I read every comment.

Subscribe

New here? Start with these:

Subscribe now

TLDR: The Link Tank turns one this month, and most of you joined along the way, so this is a proper reintroduction. I am a practising social media policy lawyer and a Continuing Professional Development (CPD) accredited LinkedIn trainer.

Before Trevisan, I lived in Switzerland, running business development for a law firm across Geneva and London. This is the story of how I worked out what LinkedIn was for, and what you can expect here now that you have arrived.

Subscribe now

Why I am writing this now

The Link Tank is a year old. In that year, it has grown from an idea to around 22,000 subscribers and Substack Bestseller status, which means the overwhelming majority of you joined somewhere along the way. You arrived mid-conversation. You have been reading my methods, my checklists, and my opinions without ever being told who is writing to you or why you should trust a word of it.

A year in feels like the right moment to fix that, so this is a proper reintroduction.

There is a second reason. Last week I wrote that every LinkedIn Substack flywheel needs an origin story, anchored in one specific moment. It would be odd to prescribe that and never publish my own. This is mine.

Every method I teach here came out of one period of my working life, when I had to win business for a firm nobody had heard of, in a market that runs entirely on reputation. If you understand that period, you will understand why I teach what I teach, and why I refuse to teach certain things at all.

Ten minutes, one story, and you will know exactly whose advice you are taking.

The years when bringing in work felt like pushing

For a long time my job was to make a law firm known to people who had never heard of it. I was living in Switzerland, director of the corporate services arm of a Panamanian firm, with business development mine to solve across Geneva and London.

The routine will be familiar to anyone who has done this work. Early flights between the two cities. Breakfast seminars where half the chairs stayed empty. A handbag full of other people’s business cards, followed by next-morning emails that went out into a silence so complete it felt personal.

The work did produce results. It produced them slowly, and in a way I could never repeat on purpose. A good introduction would arrive and I would have no idea which of the previous eighty conversations had caused it.

Was this really the best way to win work?

(I lived at the foot of the Saleve just outside Geneva)

Then I looked properly at LinkedIn

I had an account. Everybody in law and finance had an account. We treated it as somewhere to park a CV and forget the password.

What changed was that I started reading it as a map instead of a noticeboard. I could see which partners at which firms handled which kind of work. I could see which of my own contacts sat one step away from the people I needed to reach. The private client network of two cities was sitting there in plain view, and almost nobody in it was doing anything with the platform at all.

I tidied my profile so it explained what the firm actually did, in plain language, for the person checking me before a meeting. I began connecting deliberately instead of accepting whatever arrived, and sharing what I knew about the corporate services world in short pieces aimed at the people I most wanted to hear from.

The morning it worked in reverse

A few months in, an email arrived from a wealth adviser in Geneva. We had never met. He had found my profile through a mutual connection, read what I had been posting, and wanted to discuss structures for two of his clients.

When we sat down, he skipped the entire getting-to-know-you stage. He already knew how the firm worked and how I thought, because I had been showing both for months. His opening question was the one you normally reach after three meetings.

I remember walking out of that meeting and doing the arithmetic. Nobody had cold called him. Nobody had chased him. A profile and a run of posts had done, unattended, what my breakfast seminars had been failing to do at considerable expense.

Being findable, and holding up when somebody looked, turned out to matter more than any amount of pushing.

2017

The Panama Papers hit. I was made redundant while living in Switzerland. My German was non-existent as I had lived in Suisse Romande and my French was not 100% fluent, which closed most of the local job market to me. I applied for roles well below my level and heard nothing back.

What I had was a substantial network across law and finance, and a working understanding of how to make it produce business. Trevisan Consulting started in 2017 out of those two things. The idea was never complicated: show other senior professionals, particularly in professional services, how to do what I had stumbled into doing and was finding extremely effective for business development.

Other people’s version of the same story

Since then I have watched the same turn happen for the people I train.

A private client lawyer who had posted nothing in over a decade on the platform began writing short, plain explanations of the problems she solves. Referrers she had been courting for years started forwarding her pieces to clients, and the introductions began arriving with her name already trusted.

A fiduciary director who treated LinkedIn as a static business card rebuilt his profile around the questions his ideal clients actually ask. Prospects began opening first meetings by referencing things he had written, which is a very different meeting from one that opens with him explaining what a fiduciary does.

Different sectors, identical mechanism. Neither of them chased harder. Both of them became easier to find, and better to find.

What I believe about all of this

Reach is a poor measure of whether LinkedIn is working for you. I would rather you had two hundred of the right readers than twenty thousand of the wrong ones. I would also rather teach you a mechanism that keeps working than a tactic that stops working the moment the platform adjusts something.

That belief decides what appears here. You will not find growth-hacking, or screenshots of my revenue increase. You will find worked methods, and honest reporting on what the available research does and does not show; how to leverage LinkedIn’s features to generate a pipeline.

Your turn

Take a moment to think about where you are with LinkedIn right now.

Maybe you post now and then, and you want to build a presence that brings the right people to you instead. Maybe you are worn out by cold outreach and you want the warmer version of winning work, the one I stumbled into in Geneva. Maybe your profile has simply sat untouched since you last changed jobs, quietly deciding what strangers think of you.

Wherever you are starting from, the change I have described is available to you. I made it without a plan. You get to make it with one.

What you will get by following along

This is what The Link Tank exists for. Each issue takes one part of the mechanism and shows you exactly how to work it: profiles that hold up when a sceptical buyer checks you, posts that start conversations rather than collecting applause, and honest reporting on what the evidence says the platform rewards.

You will get the mechanics, and you will also get the community, which matters more. What is appropriate for a founder to say differs from what is appropriate for a lawyer or a wealth manager, and knowing the difference is most of the game. Founders, creators, consultants, employees, and people running professional services firms all read along here for exactly that reason.

If you want a taste of what that means in practice, start with the three most read issues in the archive:

1. ChatGPT Is Now Replacing Google. Here's How to Make It Recommend You shows you how to appear when clients ask Artificial Intelligence (AI) tools for experts in your field, and it began with a client telling me ChatGPT had recommended me.

2. The Smart Way to Publish on LinkedIn and Substack gives you the five-step system for publishing one piece across your website, LinkedIn, and Substack without harming your search rankings.

3. LinkedIn New Features: The Ultimate Weekly Updated Guide is the living guide I refresh with every new LinkedIn tool and algorithm change, and it is the one to bookmark rather than read once.

Ready for the first step? Subscribe below, and next week’s issue arrives with something you can apply the same day.

Subscribe now

For the few of you who would rather have help with this than a newsletter about it, my consultancy and training work lives at trevisanconsulting.com. No pitch beyond that. The newsletter is the main event here.

Over to you

I have spent this whole issue introducing myself, so it is only fair to turn it round. Introduce yourself in the comments: who you are, what you do, and the one thing you are trying to fix on LinkedIn. I will reply to every introduction, and you may well find your next useful connection in the thread below this post.

Leave a comment

You post something useful. It gets a handful of likes. A prospect asks what you charge before they’ve asked what you actually do. Somewhere else, a competitor undercuts you by fifteen per cent and wins the pitch anyway.

Or perhaps you’re earlier in the journey than that.

You’re in a demanding senior role, you know you want to build something of your own, and the only thing stopping you is the terrifying blankness of month one: no pipeline, no reputation beyond your firm’s name, no idea where the first client comes from.

Both problems have the same root. Most professionals who go independent get the order of operations backwards. They design the service, build the website, price the packages, and only then go looking for clients, discovering too late that attention takes the longest to build and should have come first.

The professionals who transition smoothly solve the audience problem while they still have a salary, so that by the time they need clients, the pipeline already exists.

This issue is about that build: the mechanism that takes you from interchangeable to only option, and the twelve-month sequence that gets you there without waiting for a slow news week or a lucky referral.

Subscribe now

Why “post consistently and hope” doesn’t fix the pricing problem

Most advice to senior professionals on building a presence stops at frequency. Post more. Show up weekly. Stay top of mind. That advice isn’t wrong, but it treats the symptom, not the cause.

The actual problem is a supply-and-demand one. When a client is choosing between several advisers, consultants or wealth managers who all describe themselves in roughly the same terms, competence stops being the differentiator. There’s nothing for the client to compare except price and availability so price becomes the deciding factor, and the professional with the deepest expertise loses to whoever quoted the lowest fee.

The problem sits in positioning, not skill. Adding more content to a crowded, undifferentiated position just increases the noise without changing the comparison the client is making.

The framework: the trust ladder

When every adviser describes themselves in roughly the same terms, expertise stops being visible, so clients compare the only thing left: price. Specificity fixes the first half of that.

Serve one clearly named audience with one clearly named problem and, in the client's mind, the field shrinks from forty options to one or two, which is what moves pricing power across the table.

Trust fixes the second half, and it's built through repeated, low-pressure contact over months rather than one well-timed pitch.

The twelve-month build, step by step

1. Narrow the positioning first. Before any content strategy, write a single sentence that says exactly who you help and with what, specific enough that a prospect reading it either says “that’s me” or moves on. Not “consultant” or “adviser,” but the outcome and the audience named together. This sentence goes in your headline and the opening line of your profile summary. If you haven’t already worked through the Headline and About section guides (see below in this article), do that first, since they cover the exact wording template this step relies on.

Simon’s profile is an excellent example of this done well.

2. Build relationships on a steady weekly cadence. Pick a realistic, sustainable number of ideal clients to connect with each week, and add a short, genuinely relevant note when you do. Consistency over the full year matters more than volume, so that by the time you’re ready to sell anything, a meaningful number of the right people already recognise your name. Keep the note free of any pitch. You’re building familiarity, not opening a sales conversation.

The practical way to find these people is LinkedIn’s Sales Navigator (I am not affiliated - a useful tool that I use with all our clients), which is a paid tool but the only reliable way to build a genuinely targeted list. Filter by industry, company headcount, geography and job title until the results match your positioning sentence, then save that list and work through it steadily each week. Expect to spend an hour or two refining the filters before the list feels right. A smaller, accurate list beats a large, loose one, because every connection you add should be someone your positioning sentence was written for.

3. Publish three types of content on rotation. Problem-solving content that names a specific, recognisable mistake your audience makes and how to fix it. Thought leadership that takes a considered position on something happening in your sector. And enough of your own perspective and personality that readers start to feel they know you, not just your job title. If you’re in a regulated sector, check anything client-facing against your compliance requirements before it goes out, particularly problem-solving content that could be read as advice.

4. Create one genuinely useful resource. A checklist, a short diagnostic, or a template that solves a real, narrow problem for your specific audience. This is what earns an email address. Send people to a simple landing page rather than asking them to message you directly, and be transparent about what happens to their data and how they can opt out, particularly if you’re contacting people based anywhere in Europe.

5. Start a short, regular note rather than a formal newsletter. People commit more easily to something that sounds low-effort than something that sounds like a subscription. A few short, useful lines sent regularly will get better engagement than an occasional long newsletter, and it keeps you present in a way that feels helpful rather than promotional.

6. Be useful in direct messages before you’re ever selling. Ask about the problem someone’s actually facing. If it isn’t something you solve, refer them to someone who can. This does more for your reputation than any post, because referrals are remembered and repaid.

7. Make your move visible when you’re ready. When you do launch, change role, or open availability, say so clearly and once, rather than hinting at it across several posts. By this point, if the previous eleven months have been done properly, you’re not introducing yourself to a cold audience. You’re confirming availability to people who already trust you.

One mechanic worth knowing: when you add a new “Experience”on your profile, LinkedIn generates a “new position” notification that reaches far more of your network than an ordinary post. Use this once, for the genuine moment you launch or open availability, and let the announcement post carry the detail. Some people trigger this notification repeatedly with cosmetic title changes to harvest attention. Resist that. Your audience notices, and the credibility you’ve spent a year building is worth more than one week of reach. Sparingly though it is an excellent way to get 1 message out to many in one go.

A worked example

Take a compliance consultant who spent a year repositioning from “I help firms with regulatory compliance” to “I help mid-sized wealth management firms prepare for their first Financial Conduct Authority (FCA) thematic review.” The narrower version felt uncomfortably small when she first wrote it.

Within the year, by consistently publishing short, specific content about thematic review preparation and building a modest list of firm principals through a preparation checklist, she stopped being one of several compliance consultants a firm might call and became the person a firm’s general counsel mentioned by name when the topic came up internally. Her day rate didn’t need to be defended. It was simply what it cost to work with her.

The objection that usually surfaces here is the fear of running out of clients. In practice, a well-chosen niche contains far more potential clients than one person can serve. When you do eventually reach the edges of it, you expand sideways into an adjacent niche rather than broadening out into everyone. A consultant who owns wealth management firms can move to private banks. One who owns legal practices can move to accountancy firms. The expertise transfers, the reputation travels with you, and you arrive in the new niche as a specialist rather than starting again as a generalist.

Where people go wrong

1. The most common mistake is copying the content style of someone successful in a completely different niche, rather than borrowing the underlying strategy. What works for a personal-brand growth expert with a broad audience will not convert for a specialist adviser, because their goals are different. One is optimising for reach. You’re optimising for a small number of the right people recognising your name.

2. The second mistake is confusing impressions with leads. A post can perform well and generate nothing, or perform modestly and bring in exactly the right enquiry. Track who’s actually engaging and whether they match your positioning, not the vanity numbers.

3. The third is trying to build every system, from pricing structures to fully built-out service pages, before a single relationship exists. The leads problem has to be solved first. Everything else can be built once there’s demand to build it for.

The copy-paste toolkit

Templates for each stage of the build. Adapt the bracketed sections to your own sector and swap the tone to match your voice, but the structure behind each one is doing the actual work.

The positioning sentence

I’m not going to repeat the formula here, since I covered it thoroughly in

and

Use those two for the exact wording template and worked examples. What this issue adds is the reasoning for why that sentence matters economically, and the twelve-month build that makes it pay off.

The connection note

Hi [name], saw you’re [specific detail from their profile or a recent post]. I work with [your specific niche] on [specific problem], so thought it made sense to connect.

Keep it under thirty words. No pitch, no link, no “let me know if I can help.” The note only needs to justify the connection.

The three content pillar prompts

1. Problem-solving: Three mistakes [your specific client type] make with [your area], and what to do instead.

2. Thought leadership: [Recent development in your sector] means [specific implication for your client type]. Here’s what I’d be doing differently.

3. Personal perspective: A short, honest line about how you work, what you got wrong once, or what you actually think about a trend in your field. This one doesn’t have a template. It just needs to sound like you.

The lead magnet landing page opener

This is a resource for [specific client type] who need to [specific outcome] without [the thing they’re afraid of or trying to avoid].

Example: For finance directors who need to prepare for an FCA thematic review without pulling their team off client work for a month.

The lead magnet creation prompt

You’re not a marketer so this is

THE BIG ONE! Copy everything in the block below into Claude or another AI assistant and it will research your sector, surface the pain points, and build the full resource with you.

I want to create a lead magnet for my professional services business. It needs to be a genuinely useful resource, a checklist, a short diagnostic, or a fill-in template, that solves one real, narrow problem for my specific audience, and that’s good enough that people will gladly exchange their email address for it. Generic advice they could find in any blog post defeats the purpose, so depth and specificity matter throughout. Before you build anything, ask me these things in one message and wait for my answers: 1. My sector and what I actually do for clients. 2. Exactly who my ideal client is: role, seniority, type and size of organisation, and country or region, since regulation and terminology vary by market. 3. The one narrow problem I want this resource to solve for them. 4. Anything I already know about what my clients complain about, ask me repeatedly, or get wrong. Once I’ve answered, research before you write. Search the web for current, credible information on my audience’s sector: the pressures they’re under right now, the regulatory or market changes affecting them this year, the language they actually use to describe this problem, and what existing free resources on this topic already cover, so mine can be more specific and more useful than what’s already out there. Tell me the three or four most important pain points your research surfaced and which one the resource will centre on, in a few lines, before you draft, then continue without waiting unless I’ve objected. Then create the fully formed resource, complete and ready to use, in British English. Choose the format (checklist, diagnostic, or template) based on which genuinely fits the problem, and tell me in one line why you chose it. Done means: a specific, outcome-led title that names my audience and the problem, so the right person instantly knows it’s for them; a short opening that shows I understand their situation, written in their language, not mine; the core content itself, meaning if a checklist, every item concrete enough to act on today, with a line on why each item matters, if a diagnostic, scored questions with an interpretation of what each score band means and what to do about it, and if a template, every section written out with instructions and a worked example, never bare headings; a closing section on what to do with their results, ending with a single, low-pressure next step that points to working with me; and length proportional to genuine usefulness, long enough to deliver a real quick win, short enough to be consumed in one sitting. Ground every factual or regulatory claim in what your research actually found rather than general knowledge, and flag anything I should verify with a compliance or legal check before publishing. Where you make an assumption about my business, state it so I can correct it. Before you finish, check the resource against this brief and fix anything that misses.

The DM referral script

That’s not something I work on directly, but I know someone who does. Want me to make an introduction?

Use this whenever a conversation surfaces a problem outside your niche. It costs nothing and it’s remembered longer than most pitches.

The launch announcement

I’m now taking on [specific number, if you want to create urgency] new clients for [specific service], starting [month]. If that’s relevant to you or someone you know, reply here or send me a message.

One post. One email to your list. Said once, not teased across several weeks.

The build cadence, at a glance

Weekly: a consistent number of relevant new connections, each with a short, specific note. Two to three pieces of content, rotating between problem-solving, thought leadership and personal perspective. A handful of genuinely helpful direct messages, with no pitch attached.

Monthly: one short, useful note sent to your growing list. A review of which content and which connections are actually leading to conversations, so you can do more of what’s working.

At month twelve, or whenever the moment is right: one clear, single announcement of your availability, sent to the list and posted once, rather than teased across several weeks.

If you want the working version of this build, I've turned the whole sequence into The Independent Consultant's Dossier: a 52-page workbook with one page per month, the full positioning worksheets, the target list method, a weekly log and 30 fully-written expert AI prompts for LinkedIn and marketing, ready to paste into Claude, ChatGPT, Perplexity or Gemini.

It's £27 and it exists for exactly the reader this issue was written for.

What’s the sentence you’d write if you had to describe exactly who you help and with what, in a way specific enough that half the room would say “that’s not me”?

Reply and tell me. I’d genuinely like to know what you land on (or DM me if you’re feeling shy🫣)

This issue draws on general market positioning and trust-economics principles rather than a single case study. The worked example above is an illustrative composite, not a real client, in keeping with client confidentiality.

Leave a comment

A LinkedIn following is applause. A Substack subscriber list is what you take home with you.

A LinkedIn following isn’t the finish line. A Substack subscriber list is where the business actually gets built.

It isn’t.

Not because they aren’t good enough. Because the follower count feels like proof, and the audit that would tell them otherwise never gets run.

A featured section that hasn’t been updated in eighteen months. A Substack About page that reads like a bio instead of a promise. A LinkedIn post from last week that mentioned “the newsletter” without a link. A profile whose two-click test takes four clicks.

None of that is fatal on its own.

What kills your pipeline is when all of it stacks up quietly, month after month, in front of an audience that would have subscribed if the doorway had been open.

The numbers behind the stall

Six to twelve months. That’s how long most LinkedIn members spend “thinking about starting a Substack” before they actually publish essay one.

A third of those who do launch a Substack abandon it within ninety days. Not because the writing was bad, but because the system underneath it was never built.

The repeat offenders: a LinkedIn headline that describes a job instead of an outcome, a featured section that leads with awards instead of the newsletter, an About page that starts with “I’m passionate about…” (😔), a Notes feed that publishes but never engages, a compliance conversation that was never had with the firm.

Almost none of this is a talent problem.

It’s a system problem you’d have fixed in an afternoon if you’d known where to look.

Two consultants get the same 5,000-follower milestone on LinkedIn.

One keeps posting daily, tells themselves growth is compounding, and looks up eighteen months later with the same 5,000 followers, no email list, and a competitor who started later now has 2,000 paid subscribers.

The other ran the audit weeks ago. The LinkedIn profile routes to a Substack that converts. The Notes routine is running. The first four essays are in draft. When the algorithm shifts, they barely notice because the flywheel is turning inside an ecosystem they own.

Same follower count. Different business.

Here’s the eight-prompt system that transforms you into the second consultant.

Prompt 1: The flywheel self-audit

Run this before you write a single Substack essay or rewrite a single LinkedIn post. It’s the same first pass a LinkedIn strategist would run on your account, pointed at your own setup while you still have time to fix what it finds.

💬 Copy this prompt into Claude

I'm a LinkedIn creator building toward an owned Substack audience.

Current LinkedIn following: [NUMBER]
Current Substack subscribers: [NUMBER OR "0"]
Niche: [ONE-SENTENCE DESCRIPTION OF WHO YOU SERVE]
Goal: [SPECIFIC — e.g. "reach 1,000 free Substack subs in 90 days"]

Act as a LinkedIn–Substack flywheel strategist doing a first-pass
structural review before I commit to a publishing cadence. Based on
standard flywheel practice at my stage, walk through these five
categories and tell me, for each one, what a strategist would ask
for and what commonly goes wrong:

1. Discovery — LinkedIn side: headline, featured section, profile
   copy, and the two-click path from any post to my Substack
   subscribe box
2. Depth — Substack side: About page, welcome email, cadence,
   post structure, call-to-action pattern
3. Connective tissue — Notes: publishing frequency, engagement
   pattern, and whether I've built a home network of adjacent
   writers before expecting reach
4. Positioning — do my LinkedIn headline, Substack About page,
   and pinned Note all promise the same specific outcome to the
   same specific reader?
5. Compliance and platform risk — for regulated professionals,
   what am I posting on LinkedIn that I would need to pre-clear
   on Substack, and where does that conversation need to happen
   with my firm?

For each category, ask me direct yes/no questions that would
surface a real gap, not generic checklist items. Where I answer
"I'm not sure," flag that as a priority to fix this week, not
later.

💡 Why this works: “Audit my LinkedIn and Substack” produces a generic checklist. This produces the specific, categorised gap list a strategist would actually build, and it puts your uncertain answers at the top instead of the bottom.

If this is useful, it’s worth knowing it’s one piece of a bigger system.

The prompt above assumes you’ve already committed to running LinkedIn and Substack as one flywheel, not two disconnected accounts. That’s the strategy. What follows is the execution. I’ve built the same kind of copy-paste system for every stage that comes before this one:

• How to Write a LinkedIn Headline That Converts Profile Views into Enquiries — the full guide for rewriting a job-title headline into one that routes strangers to your Substack before you touch anything else

• Is Substack Worth It for Professionals? 10 Questions I Get Asked Every Week — the honest answers to the questions every regulated professional asks before they publish essay one

• From Invisible to Inevitable: Inside the Allocator’s Browser Before the First Meeting — how a serious buyer actually researches you before they book the call, and what your LinkedIn and Substack need to look like when they do

• The LinkedIn–Substack Flywheel — the two-platform system that turns rented reach into an audience you own

• LinkedIn Switched Off My Account Overnight. Here’s What I’d Do Differently. — the case study for why waiting to build your owned audience is the biggest risk in your business

Prompt 2: The LinkedIn profile teardown

The highest-frequency failure point in the entire flywheel. Almost always an accumulation of small drift, not a single mistake.

A headline written in 2022 for a job you’ve since evolved past. A featured section that still leads with a webinar from last year. An About paragraph that opens with credentials instead of a promise. A pinned post that no longer represents what you write about.

None of it wrong on its own. All of it reads as unfocused when a stranger lands on your profile and has three seconds to decide whether to subscribe.

💬 Copy this prompt into Claude

Here is my current LinkedIn profile: [PASTE HEADLINE, ABOUT
SECTION, FEATURED SECTION CONTENTS, AND YOUR THREE MOST RECENT
POSTS]

Here is my Substack and what I write about:
[PASTE SUBSTACK NAME, ABOUT PAGE, AND YOUR LAST TWO POSTS]

Walk through my LinkedIn profile section by section and check:
1. Does my headline name a specific outcome I deliver for a
   specific reader, or does it describe a job title?
2. Does my featured section route a stranger to my Substack
   in the first slot, or is my Substack buried behind awards,
   press, or old lead magnets?
3. Does the first three lines of my About paragraph tell a
   stranger who this profile is for and what they will get if
   they follow me, or does it open with credentials?
4. Do my last three posts each contain either an explicit
   Substack pointer or content that would make a stranger want
   to subscribe?
5. If I opened this profile in an incognito window as a
   stranger who just clicked one of my posts, could I reach
   the Substack subscribe box in two clicks?

For anything that isn't routing correctly, tell me exactly what
to rewrite and give me a specific replacement, in my voice,
that I can paste in today.

⚡ Run this the same week you commit to publishing weekly on Substack, not after. A profile fix takes an afternoon when the audience is 5,000.

It takes a lot longer to explain the pivot when the audience is 50,000 and the old positioning is what they followed you for.

🔒 Two prompts down. Six to go before your flywheel is actually turning.

The self-audit and the profile teardown tell you where the gaps are but they don’t close the ones underneath the ones you can see.

Paid subscribers get the next six prompts in full, plus the tools that go with them:

✓ Prompt 3: The Substack About page rewrite - takes whatever bio you’re currently using and turns it into a specific promise a stranger will subscribe to, using the “This is for [X]. Every [cadence] I send [format] on [topic]. Subscribe if you want [outcome]” structure that converts.

✓ Prompt 4: The four-essay launch sequence generator - drafts the exact opening arc every flywheel Substack needs: manifesto, origin story, contrarian take, and playbook. Personalised to your niche, in your voice, ready to schedule.

✓ Prompt 5: The 15-minute Notes routine builder - takes your existing LinkedIn hooks and turns them into a week’s worth of Substack Notes across three formats (quote-plus-take, mini-list, question), plus the daily engagement pattern that pulls the flywheel into motion.

✓ Prompt 6: The LinkedIn-to-Substack pointer prompt - writes the specific one-line pointers you can drop into any LinkedIn post to route readers to your Substack without triggering the algorithm penalty for external links.

✓ Prompt 7: The compliance conversation script - the exact two-sentence reframe to take to your firm’s compliance team, plus the pre-clear versus publish-freely matrix that turns a “can I?” conversation into a “here’s how I will” conversation.

✓ Prompt 8: The 30-day flywheel review - runs a structured retrospective on your first month’s data (LinkedIn engagement, Substack conversion, Notes reach) and tells you exactly which of the three moves to double down on and which to fix.

Paid subscribers also get:

Got it. This is a much better offer than “60+ prompts” — it’s a real, delivered, priced product with specific numbers. Here’s the replacement teaser block.

Paid subscribers also get The Instant Content Kit for LinkedIn + Substack (2026) — free.

The same kit that sells on Gumroad, ready-to-use, delivered instantly, yours forever which includes:

✓ The 2026 Viral Content Calendar - every major trend, social media holiday, and content prompt mapped across the full year, so you never miss a moment worth posting into

✓ 150 Scroll-Stopping Hooks — openers built to grab attention on Substack Notes and LinkedIn posts. Pick one, tailor it to your topic, publish

✓ 93 High-Impact Calls to Action — the specific closes that get people to click, comment, reply, or share, in the register a professional audience actually responds to

✓ The Link Tank™️ Content Library Creator GPT — a custom GPT that turns your niche into 30 days of content in minutes, not hours

Sold separately, this kit is a full product on Gumroad. Paid subscribers here get it as part of the subscription, alongside the eight prompts above and every Sunday deep-dive to come.

Read more

It comes from lawyers with 15,000 LinkedIn followers who have realised, uncomfortably, that they do not own a single one of them. It comes from financial advisers who spent three years building a personal brand on a platform that can throttle their reach tomorrow without explanation. It comes from consultants who watched a competitor lose an account overnight because a compliance officer decided LinkedIn was “too public”.

The panic is not really about LinkedIn.

For the last decade, the advice for professionals was straightforward: build on LinkedIn. Post consistently. Grow your network. That advice was correct. It is now incomplete.

In 2026, the operators who are quietly compounding - the ones getting inbound work while their peers get impressions - are running a two-platform system. LinkedIn does one job. Substack does another. Notes stitches them together.

This is the flywheel. And once you see it, you cannot unsee it.

By the end of this essay you will have a scorecard to check your own setup, a four-essay sequence to launch your Substack, a Notes routine you can start tomorrow, and two sentences to take to your compliance team on Monday.

Subscribe now

The problem with one-platform strategies

Let me put a number on the risk.

Substack’s own data team, after studying thousands of publications, recommends one long-form post per week as the growth benchmark. Their reasoning is simple: growth is tied to output, and one deep piece a week is the minimum viable rhythm for compounding.

LinkedIn’s advice, in practice, is almost the opposite. Post daily. Comment daily. Feed the algorithm. Miss a week and your reach collapses.

Both platforms are telling the truth about themselves. But they are telling you two different truths, and if you only listen to one, you end up with one of two failure modes:

The LinkedIn-only creator has reach but no ownership. Every post is rented. Every follower is a lead they cannot email. Every algorithm change is an existential event.

The Substack-only creator has ownership but no discovery. The essays are beautiful. The list grows by five subscribers a week. The compounding never starts because nobody new ever arrives.

Neither of these is a business. They are hobbies with different risk profiles.

What the flywheel actually is

The flywheel is not a content calendar. It is a division of labour between two platforms that were, quietly, built to work together.

LinkedIn is your discovery engine. It is where new people find you. Its job is reach, not depth. You post to be seen by strangers.

Substack is your owned audience. It is where interested people go deeper. Its job is depth and ownership, not reach. You publish to be read by people who already trust you enough to give you their email address.

Notes is the connective tissue. Substack Notes now operates on the same discovery mechanics that made early LinkedIn work — engagement-based distribution, in a feed of people already primed to subscribe rather than scroll.

Here is what the loop looks like in practice:

1. A stranger sees your LinkedIn post → engages → visits your profile

2. Your profile links to your Substack → they subscribe (free)

3. They read your Sunday long-form → trust deepens

4. They see your Notes → they restack, share, and pull in their network

5. Their network subscribes to your Substack → the loop restarts, now inside Substack’s ecosystem

6. A percentage of subscribers convert to paid, to your service, or to your product

Every rotation of the wheel deposits owned subscribers into an asset you control.

This is what Lenny Rachitsky built to over one million subscribers, publishing once a week. It is what Nicolas Cole built publishing twice a week. And in a niche much closer to yours, it is what Kristina God — my co-host for the August bootcamp - used to reach 22,000+ subscribers and Substack’s Top 15 International Bestseller list.

None of them are on LinkedIn all day. All of them own their audience.

The Flywheel Scorecard

Before you change anything, find out where you actually stand.

How to complete it

Set aside 10 minutes. Open your LinkedIn profile and your Substack in an incognito or private browser window - this is the single most important instruction on the page. You are checking how a stranger sees you, not how you see yourself.

Go through each statement in order. Tick Yes only if the answer is unambiguously true today, not “mostly”, not “I’ve been meaning to”, not “next week”. Tick No for everything else. Ambiguity is a No.

When you are done:

1. Total your Yes answers and read the scoring rubric below.

2. Date the scorecard — today’s date, and a retake date 30 days from now. The value of this exercise is the delta between the two.

3. Fix your No answers in order — discovery first (1–4), then depth (5–7), then Notes (8–10). Do not skip ahead. A broken doorway makes every downstream move less effective.

Download printable version

The 10 statements

Discovery (LinkedIn side)

1. My LinkedIn headline names the outcome I deliver, not just my job title. ☐ Yes ☐ No

2. My Substack link appears in my LinkedIn featured section, first slot. ☐ Yes ☐ No

3. A stranger can get from any of my LinkedIn posts to my Substack subscribe box in two clicks. ☐ Yes ☐ No

4. At least one of my last five LinkedIn posts explicitly pointed to my Substack. ☐ Yes ☐ No

Depth (Substack side)

5. I published a long-form Substack post in the last 7 days. ☐ Yes ☐ No

6. My Substack About page names who it is for and what they will get, in the first three lines. ☐ Yes ☐ No

7. My most recent post ends with a single, specific call to action, not a generic “subscribe”. ☐ Yes ☐ No

Connective tissue (Notes)

8. I published at least one Note in the last 3 days. ☐ Yes ☐ No

9. I restacked or replied to at least three other writers’ Notes this week. ☐ Yes ☐ No

10. I could name three Substack writers in an adjacent niche who would notice if I posted something worth restacking. ☐ Yes ☐ No

How to read your score

8 or more: Your flywheel is turning. Focus on rhythm and rest, not more tactics.

4 to 7: You have a wheel, but it is stuck. Fix the No answers in order — discovery, depth, then Notes.

0 to 3: You have a hobby, not a system. Start with move one below and do nothing else for two weeks.

Screenshot this. Send it to a colleague. Come back to it in 30 days.

Move one: turn LinkedIn into a doorway, not a destination

Most LinkedIn profiles are designed to impress. Flywheel profiles are designed to route. The job of your profile is not to prove you are credible - your posts do that. The job is to move an interested stranger to your Substack in two clicks.

The headline formula

[Outcome I deliver] · [Who I deliver it for] · [One proof point] · [Owned-audience hook]

Filled in from my own profile:

Accelerating Visibility & Revenue · CPD Accredited LinkedIn Training for Regulated Professionals · 5 Citywealth Awards · Sunday essays at The Link Tank

The last clause is the doorway. It tells every stranger who reads your headline that there is somewhere else to go.

The featured section order

Rearrange your featured section into exactly this sequence:

1. Slot 1: Your Substack — thumbnail, one-line pitch, direct link

2. Slot 2: Your free lead magnet — a guide, checklist, or template gated by email

3. Slot 3: Your paid service or product — kept last on purpose, so it feels earned

Nothing else. Remove awards, press mentions, and old webinars. They belong on the profile body, not the featured section.

The two-click test

Open your LinkedIn profile in an incognito window. Time yourself. Can a stranger reach your Substack subscribe box in two clicks or fewer? If it takes three, your doorway is broken. Most commonly, the problem is a featured tile that opens a preview instead of linking straight through - fix the link target, not the design.

Move two: the first four Substack essays

Do not think in terms of “publish weekly”. Think in terms of a sequence. These are the first four posts every flywheel Substack needs, in order. Together they take a stranger from “who is this?” to “I want to work with this person”.

Post 1 — The Manifesto
What you believe about your field that most people get wrong.
Prompt: “The industry believes X. I believe Y. Here is why Y is right, with three examples.”

Post 2 — The Origin Story
Why you, why now. Anchored in one specific client moment.
Prompt: “The moment I realised [current work] mattered was when [specific client, specific problem, specific outcome].”

Post 3 — The Contrarian Take
A widely held belief in your niche that you disagree with. This is the one people will restack.
Prompt: “Everyone in [niche] tells you to do X. I think X is wrong for [audience]. Here is what to do instead.”

Post 4 — The Playbook
The how-to your audience will actually implement. Concrete, screenshottable, complete.
Prompt: “The exact steps I would take if I were starting [thing] from scratch tomorrow.”

Write post one this Sunday. Schedule post two for next Sunday. You now have a two-week runway before you have to think about content again.

Move three: the 15-minute Notes routine

Notes is the cheapest, fastest lever on Substack. Most people ignore it because it feels like Twitter. That is exactly why it works - the people who show up consistently get disproportionate reach.

Run this routine every weekday. Set a 15-minute timer.

Minutes 0–5: Restack two Notes with a one-sentence add.
Not “great post”. Add a genuine observation, a counter-point, or a specific example. Restacking with a comment is worth ten times a silent restack because the algorithm treats it as original content.

Minutes 5–10: Publish one original Note.
Use one of these three formats. Rotate them across the week so your Notes feed does not look monotonous.

Format A — Quote plus take (best for restacks):

A following is not an audience. A platform is not a business.
Most creators still cannot tell the difference.

Format B — Mini-list (best for saves):

Three things I would do differently if I started my Substack today:

1. Publish the manifesto first, not last.

2. Link the Substack from LinkedIn on day one.

3. Restack before I posted.

Format C — Question (best for replies):

Genuine question for anyone with a Substack over 5k subscribers: what percentage of your growth came from LinkedIn versus Notes?

Minutes 10–15: Reply to comments on your most recent long-form post.
Every reply is a small trust deposit and a nudge to the algorithm that your post is still alive.

Fifteen minutes. Five days a week. That is the whole routine.

I use Writestack to save my sanity and schedule my notes in advance.

For regulated professionals: the compliance mini-toolkit

If you are under an FCA, SRA, or firm-level social media policy, here is how to make the case internally and stay out of trouble.

The two sentences to take to compliance

“Substack is a publishing platform, not a social network. My posts will follow the same review and disclosure process as any bylined article I write for a trade publication.”

That single reframe changes the conversation from whether you can publish to how. Most compliance teams have an existing article-review workflow you can slot into.

What to pre-clear versus publish freely

• Pre-clear: anything referencing specific clients, specific transactions, regulated products, performance figures, forward-looking statements, or firm strategy.

• Publish freely (within firm policy): commentary on public developments, generic educational content, personal opinion clearly framed as such, book and article recommendations, event write-ups.

The three topics to avoid regardless of platform

1. Client-identifying details, even in composite or “friend of a friend” framing.

2. Predictions about markets, rates, or regulatory outcomes stated as fact.

3. Anything that could be read as promotion of a specific regulated product without the required disclosures.

Print this section. Put it in front of your compliance officer. You will have a faster conversation than you expect.

Three common mistakes (and the specific fix for each)

Mistake 1: “I moved my LinkedIn audience to Substack and only 3% subscribed.”
Why: your About page is a bio, not a promise. Strangers do not subscribe to a person. They subscribe to a specific, repeated benefit.

Fix: rewrite the first three lines of your About page in this shape: “This is for [specific person]. Every [cadence] I send [specific format] on [specific topic]. If you want [specific outcome], subscribe below.” Watch the conversion rate double within a fortnight.

Mistake 2: “My Substack posts get 40 opens and no restacks.”
Why: your CTA is “subscribe” or “share”. Both are generic. Neither gives the reader a reason to act now.

Fix: replace the closing CTA with one specific ask tied to the post’s argument. “If you disagreed with move two, restack this and tell me why.” Specific asks get 5 to 10 times the action rate of generic ones.

Mistake 3: “I post daily Notes and nothing happens.”
Why: you are publishing before you are engaging. Notes rewards reciprocity - the algorithm surfaces you to the networks of people you have already engaged with.

Fix: reverse the routine. Restack two Notes and reply to three writers before you publish your own. Do this for two weeks and measure the shift.

The uncomfortable question

The reason most people will not build a flywheel is not that they do not understand it but that starting a Substack in 2026, when you already have a LinkedIn following, feels like starting from zero. A proverbial mountain to climb.

The question is whether you move them to something you own, before the platform you rented decides you cannot.

What to do this week

• Today: take the scorecard. Screenshot your result.

• Sunday evening: rewrite your LinkedIn headline and featured section using move one.

• Monday: draft essay one (the manifesto) using the prompt in move two.

• Tuesday to Friday: run the 15-minute Notes routine.

• Next Sunday: publish essay one.

If your scorecard came in under 7, and you want the full system laid out step by step - the profile teardown, the About page rewrite, the four-essay sequence, and the Notes cadence - Kristina God and I are running a three-day bootcamp on 18–20 August that walks you through the whole flywheel.

Join The Waiting List Here : linkedin-substack-flywheel-bootcamp.pplx.app

Next Sunday, I will go into more depth on move one: the LinkedIn profile teardown, with a live before-and-after. If you want that in your inbox, subscribe below.

The wheel does not turn itself, but once you start it, it does not stop (in the best possible way).

Share The Link Tank™️

Hi I’m Melanie, a LinkedIn consultant & strategist who turns insider knowledge and sharp strategy into measurable results

TL;DR

• How to decide who you actually want to see in your LinkedIn feed

• How to remove noise fast using the three‑dot menu and unfollow / “Not interested”

• How to train the algorithm with intentional comments, saves, follows, and DMs

• How to use a copy‑and‑paste AI prompt with your LinkedIn data export to identify high‑priority connections you have never messaged or have let go cold

LinkedIn can either feel like a useful professional tool or an exhausting stream of noise. The difference is rarely the platform itself; it is usually whether the feed has been curated with intention.

A well-curated LinkedIn feed can become a daily source of insight, client opportunities, warm relationships, market intelligence, and better content ideas. A poorly curated one tends to fill with generic motivation, irrelevant updates, engagement bait, and people who have nothing to do with current goals.

LinkedIn gives users meaningful control over what appears in the feed through following, unfollowing, commenting, saving, profile visits, messaging, and content-hiding actions.

The good news is that feed curation is not complicated. It is simply a matter of deciding who matters, removing what does not, and repeatedly sending LinkedIn stronger signals about the people and topics worth prioritising.

Subscribe now

Why feed curation matters

Most people underestimate how much their LinkedIn feed shapes their thinking, behaviour, and opportunities. What appears in the feed affects who gets noticed, whose ideas feel familiar, which conversations are easy to join, and which prospects stay top of mind.

A curated feed can help with:

• Seeing ideal clients and prospects more often, which creates more natural opportunities to engage.

• Staying close to referral partners, peers, collaborators, and industry voices worth knowing.

• Spotting trends, updates, launches, and market shifts earlier.

• Finding better inspiration for posts, comments, newsletters, and conversations.

• Reducing cognitive clutter so time on LinkedIn feels productive rather than draining.

This is why feed curation is an attention-management tactic.

Start by defining who you actually want to see

Before changing settings or cleaning up the feed, it helps to decide what “valuable people” means.

For one person, that may mean potential clients. For another, it may mean industry experts, collaborators, decision-makers, journalists, employers, founders, or peers. The point is not to follow the most popular people. The point is to follow the people most aligned with present goals.

A useful way to structure this is to build a simple “must‑see” list with categories such as:

• Potential clients

• Referral partners

• Industry experts

• Peers and community

• Media and commentators

• Companies and organisations worth tracking

Many LinkedIn practitioners recommend deliberately following a focused set of relevant people rather than trying to consume content from everyone in a large network.

Following first and connecting later can also help keep the feed more intentional, because it separates feed quality from network size.

Clean up the existing feed

Once a clearer target is in place, the next step is to remove noise. This does not require disconnecting from large numbers of people. In most cases, unfollowing is enough.

Practical ways to clean up the feed include:

• Unfollowing people whose content is repetitive, irrelevant, or low quality while remaining connected.

• Using “Not interested” or “Hide post” whenever unwanted content appears, which helps retrain the feed over time.

• Reporting obvious spam so it is less likely to keep resurfacing.

• Resetting feed preferences periodically if the signal has become too messy over time.

These kinds of searches ensure your feed fills with people and firms that are commercially relevant, not just broadly “interesting”.

Where to click (so nothing is left to guesswork)

LinkedIn gives control, but much of it is hidden behind small interface elements. Knowing exactly where to act removes friction.

To unfollow someone but stay connected:

• Click the three dots (- - - ) on their post

• Select “Unfollow [Name]”

• Or go to their profile → click “Following” → toggle to “Unfollow”

To reduce unwanted content:

• Click (- - - ) on a post → select “Not interested in this post”

• Repeat this consistently for anything irrelevant

To prioritise someone’s content:

• Go to their profile

• Click the bell icon (top right of profile header)

• Select “All posts” or “Top posts”

To follow without connecting:

• Use the search bar

• Click “Follow” instead of “Connect”

.

Train the algorithm with your behaviour

The feed is shaped by behaviour as much as by settings. Every reaction, comment, save, profile visit, message, and moment of attention teaches LinkedIn what deserves more space.linkedin+1

Some signals appear to be stronger than others. Commenting meaningfully tends to be a stronger signal than liking, and saving a post can also indicate that the content is especially valuable.linkedin+1 Dwell time matters too: spending time reading a post tells LinkedIn the content is relevant, even without a visible reaction.

This means feed curation is not only about removing bad signals. It is also about intensifying the good ones.

Helpful habits include:

• Commenting on posts from people worth seeing more often.

• Saving genuinely useful posts rather than scrolling past them.

• Visiting the profiles of important people so LinkedIn recognises that interest.

• Sending occasional direct messages to strengthen relationship signals.

• Avoiding “hate engagement”, because arguing with or reacting to poor content still trains the algorithm to show similar material.

A simple rule works well here: engage with what should expand and ignore what should disappear.

Use search deliberately (not passively)

Search is one of the most underused tools for feed curation. Instead of waiting for the algorithm to surface the right people, you can actively define who appears.

Try searches such as:

• “General Counsel London”

• “Head of Compliance UK Financial Services”

• “Family Office Principal Europe”

• “Private Client Partner Law Firm”

Then:

• Click “People”

• Filter by location, industry, or current company

• Follow first, observe content, then decide whether to connect later

This approach keeps your feed aligned with commercial relevance rather than network size.

How to comment so LinkedIn shows you more of the right content

Commenting is one of the strongest signals you can send, but only if done with intent.

A simple structure works well:

• Acknowledge the point

• Add perspective or experience

• Extend the idea

For example:
“This is a strong point on employee advocacy. In regulated firms, I often see hesitation due to compliance concerns. One approach that works well is creating pre-approved content frameworks rather than restricting activity entirely.”

This type of comment does three things at once:

• Signals to LinkedIn that this content matters

• Strengthens your visibility with the right audience

• Positions you through contribution rather than broadcasting

Short, generic comments do not have the same effect.

What you should notice within a week

Feed curation is not instant, but it is responsive. Within a few days of consistent action, most people begin to see measurable changes.

You should start to notice:

• More posts from people you have recently engaged with

• Fewer irrelevant or repetitive updates

• Increased visibility of your target audience (clients, peers, partners)

• More useful ideas for your own content

If this does not happen, it usually points to inconsistent signals - most commonly engaging with content that is not actually relevant.

A 10-minute feed reset (for immediate impact)

If you want a simple way to track this without building your own system from scratch, I have created the LinkedIn Feed Curation Kit – a ready‑to‑use toolkit with:

• A must‑see list template

• A feed clean‑up tracker

• Daily and weekly checklists

• A relationship re‑engagement watchlist

• LinkedIn Feed re-engagement playbook

• The copy‑and‑paste AI prompt from this article, pre‑formatted
  

  You can access it here: LinkedIn Feed Curation Kit on Gumroad.

  

  Access Kit

Add better signals, not just fewer bad ones

Curating a feed is not only subtraction. It is also about seeding the feed with better sources.

There are several ways to do that well:

• Use LinkedIn search to find people by role, industry, location, company type, or topic, then follow them.

• Follow company pages, especially clients, prospects, competitors, and relevant publications.

• Join groups connected to real interests or industries so more niche discussions begin surfacing in the main feed.

• Subscribe to useful newsletters so higher-quality long-form ideas enter the content mix

• Turn on notifications for a small number of top-priority people whose posts should never be missed.

This is where LinkedIn starts becoming more than a scrolling app. It becomes a personalised briefing document built around chosen priorities.

The 2026 reality on hashtags

Hashtags are still usable on LinkedIn in 2026, but they are no longer a meaningful reach lever on their own. Recent commentary across LinkedIn and social media analysis consistently describes hashtags as supporting signals rather than growth engines.

That distinction matters. In earlier years, many users treated hashtags as if they could materially boost distribution simply by being added to a post. Current guidance suggests that LinkedIn now relies much more heavily on topical clarity, natural keyword use, engagement quality, and broader relevance signals than on hashtag use alone.

In practical terms, hashtags now help most with:

• Clarifying the topic of a post.

• Supporting niche discovery in smaller communities.

• Adding light search and categorisation value.

What they do not do is rescue weak content or create reach through volume. Posts are more likely to perform because they are clear, timely, relevant, and engaging than because they contain a long block of tags.

A sensible 2026 approach is:

• Use 3 to 5 relevant hashtags at most.

• Prefer specific and contextual hashtags over broad generic ones.

• Avoid stuffing posts with long lists of tags.

• Focus more on keyword-rich opening lines and clear subject matter than on hashtag strategy.

In other words, hashtags are best treated as labels, not levers.

Turn your archive into a relationship map

Curating your feed is not only about what you see today; it is also about who you have quietly ignored or never followed up with.

One powerful way to align your feed with your goals is to analyse your LinkedIn data export and surface connections you have never actually spoken to, or have allowed to go cold. You can then use the resulting spreadsheet as a practical re-engagement list and, combined with your curated feed, focus your daily comments and messages on people who actually matter to your current goals.

You can give this prompt to your preferred AI tool after uploading your LinkedIn data export (specifically Connections.csv and Messages.csv):

Copy–paste prompt:

What a well-curated feed really gives you

The real payoff is not just a nicer experience. It is better professional proximity.

When the right people show up regularly in the feed, it becomes easier to comment, easier to stay visible, easier to notice opportunities, and easier to create content that is grounded in real conversations. That can lead to warmer outreach, stronger relationships, sharper positioning, and more relevant ideas.

The feed does not need to be perfect. It simply needs to be pointed in the right direction often enough that LinkedIn starts working with intention rather than by accident.

A curated feed is not about seeing more content. It is about seeing more of the right content.

“Melanie, be honest… is Substack actually worth it for someone like me?”

Then the floodgates open.
Will I get real clients from it?
Do I have time for yet another platform?
What about compliance and my firm’s social media policy?

I’ve now answered the same 10 questions so many times that it’s become a playbook.
This issue is that playbook.

Subscribe now

1. “But do you actually get clients from Substack?”

Short answer: yes – but not in the “post once, wake up to 47 new enquiries” way people secretly hope for. It works because Substack gives people a quiet, long‑form way to get to know how you think before they ever book a call.

What actually happens is this: someone finds me on LinkedIn, they like a post or save a carousel, then they join Substack “just to see what I’m about.” Over a few weeks they read a couple of issues, hit reply to one of them, and that email reply turns into, “Could we have a quick call about our team?” That’s the real client journey – social for discovery, Substack for depth, then a conversation.

If you want Substack to generate work, you have to treat it as part of your client pathway, not a separate creative hobby. End issues with a clear, low‑pressure next step (hit reply, download a guide, join a webinar, book a diagnostic) and then track who quietly moves from reader to client.

2. “Are LinkedIn and Substack competing platforms – or a power duo?”

They’re a power duo. LinkedIn is your shop window; Substack is your private members’ club. One is where people discover you; the other is where they decide whether to trust you.

On LinkedIn, your content is competing for attention in a busy feed, alongside job updates, humblebrags, and memes. Substack drops you straight into someone’s inbox or reading app with no algorithm in the middle. For professionals, this is gold: you can unpack a topic properly, add nuance, and speak to the real concerns your clients would never put in a public comment.

The magic happens when you make them work together intentionally: LinkedIn posts seed ideas and send people to subscribe, and Substack issues send people back to connect on LinkedIn, attend events, or share your work inside their organisation. That’s when the flywheel starts to turn.

3. “What should I put on Substack that I don’t already post on LinkedIn?”

Think of LinkedIn as the trailer and Substack as the full film. LinkedIn gets the short, sharp, scroll‑stopping version; Substack gets the detailed, contextual, sometimes slightly less “polished” truth.

On Substack, you can share:

• Deeper case studies (with anonymisation where needed)

• Practical frameworks with step‑by‑step breakdowns

• “Behind the scenes” of how you think, work and decide

• Longer, more reflective pieces that simply don’t fit a feed

You can also write for a narrower audience than you do on LinkedIn. For example, on LinkedIn you might say “professional services”, while on Substack you write directly for “heads of marketing at boutique law firms who are trying to get partners to post.” The more specific you are, the more people feel, “this is literally about me.”

Download Guide

4. “How often do I need to publish on Substack (without losing my mind)?”

You don’t need a heroic publishing schedule. For most professionals, once a week is realistic and more than enough to build a strong presence over time.

Here’s a cadence that works for my clients:

• One main issue per week (your “flagship” article)

• Optional: one shorter note or update when you have something timely to say

• A simple repurposing habit where each issue becomes 2–4 LinkedIn posts

The goal is consistency, not volume. I’d rather see you publish 40 thoughtful issues in a year than sprint for six weeks, burn out and disappear. Set a schedule that respects your client work, add buffer weeks into your content plan, and accept that some issues will be “solid and useful” rather than “my magnum opus.” That’s fine.

5. “Isn’t Substack totally saturated – is there any point starting now?”

It can feel saturated because you’re suddenly aware of how many writers are already there. But most of your peers and competitors are still doing… nothing. They’re lurking, thinking about launching “one day,” or treating Substack as a place to occasionally repost old blogs.

What matters isn’t whether there are other newsletters in “law” or “finance” – there are. What matters is whether there’s anyone speaking clearly to your niche, in your tone, with your mix of expertise and pragmatism. There usually isn’t. Saturation is rarely the real issue; fear of being visible is.

If you bring specificity (who you’re for), point of view (what you believe), and consistency (showing up when you say you will), you will stand out – even in a busy category. Professionals don’t need 100 newsletters. They need one they trust and actually read.

6. “How big does my LinkedIn audience need to be before Substack is worth it?”

You don’t need a huge audience. You need a relevant one. I’ve seen people with a few hundred LinkedIn connections build a meaningful Substack list because every single subscriber sits firmly inside their target market.

Substack starts to feel powerful well before you hit “big” numbers. Imagine you have 150‑300 subscribers, mostly existing or ideal clients. If 40–50% open each issue, a handful click through and one or two hit reply, that’s already better attention than a typical feed post that briefly flashes past thousands.

So don’t wait until you have X followers to “justify” a newsletter. Start now, tell your existing network what you’re doing, and grow steadily. Your early issues are where you find your voice and refine your focus. By the time your audience is larger, you’ll know exactly what they value.

7. “How do I promote my Substack on LinkedIn without looking spammy?”

The key is to integrate, not interrupt. Promotion shouldn’t feel like a hard sell; it should feel like the natural next step for anyone who enjoyed your content.

A few practical moves that don’t trip compliance alarms:

• Put your Substack link in your LinkedIn “Contact info” and “Featured” section

• Mention it once in your About section as the place to get your deeper insights

• Occasionally share a short excerpt on LinkedIn with a line like:
  “This is an excerpt from today’s issue of my newsletter for [audience]. If you’d like the full breakdown in your inbox, you can join it here.”

For private outreach, think targeted, not mass. A personalised note to someone who regularly engages with your posts is very different from a blanket “join my newsletter” message to 500 people. Use language that makes it feel like an invitation, not a transaction.

8. “Should I charge from day one, or start with a free newsletter?”

For most professionals, I recommend starting free and treating Substack initially as a trust‑building and pipeline‑warming tool, not your primary revenue line. Once you’ve proven that the newsletter is valuable and sustainable, you can decide whether to add a paid tier.

Remember: your main “product” is likely your advisory, training, or services. Substack is part of the marketing and relationship‑building layer that feeds those. A paid tier can make sense later for:

• Deep‑dive workshops or office‑hours style calls

• Premium resources (templates, checklists, in‑depth guides)

• A more intimate community for highly engaged clients and peers

You don’t need to figure out your entire monetisation strategy before you write Issue 1. Start with free, get momentum, and then make an informed decision rather than a theoretical one.

9. “Which metrics actually matter on Substack (compared with LinkedIn)?”

On LinkedIn, people obsess over views, impressions and likes. On Substack, the metrics that matter are quieter but more meaningful: subscribers, open rates, click‑throughs and, crucially, replies.

I’d start by watching:

• Subscriber growth over time (is it trending in the right direction?)

• Open rate (are people actually reading, or just signing up and forgetting?)

• Click‑through rate (when you include a link, does anyone care enough to click?)

• Replies and forwards (who is treating this as valuable enough to respond or share?)

The most important thing is to connect those numbers back to real outcomes. Did a certain issue lead to more LinkedIn messages? Did a CTA drive sign‑ups to a webinar or downloads of a guide? Track that. Otherwise it’s easy to swap one set of vanity metrics for another.

10. “How do I stay consistent on Substack when I’m already busy on LinkedIn?”

You build a system, not a heroic streak. Consistency comes from simplifying decisions, batching where possible, and repurposing smartly between platforms.

Here’s a simple model my clients like:

• Choose one or two core content themes for the next quarter

• Block a recurring slot in your calendar to draft each issue (even 60–90 minutes)

• Turn each Substack into several smaller LinkedIn posts, rather than starting from zero

Also, give yourself permission to have “lighter” issues: curated links, a short reflection, a Q&A. Your audience doesn’t expect a thesis every week. They expect you to show up, be useful and be honest. If you can do that consistently, you’re already ahead of almost everyone who “means to start a newsletter someday.”

Your first 30 days: a plan you can actually follow

Most people stall at the same point. They accept that Substack is worth it, then freeze at “so what do I actually do first?” Here is the plan I give clients who want to be live within a month without derailing their client work.

Week one: decide who it is for and what they get.

Write one sentence before you write anything else. “This is a newsletter for [specific person] who wants to [specific outcome].” For example, “heads of marketing at boutique law firms who want their partners to post without it becoming a compliance headache.” Resist the urge to widen it. The narrower your sentence, the easier every future issue becomes to write, because you always know exactly who is reading.

Week two: set the publication up properly.

Work through four things:

1. Name the publication after the value it delivers rather than after yourself. A reader deciding whether to subscribe wants to know what they get, not who you are.

2. Write a short About page that says who it is for, what they will learn, and how often you publish.

3. Write a one-paragraph welcome post that new subscribers see the moment they join.

4. Add your Substack link to your LinkedIn Featured section and Contact info, so discovery flows straight into depth.

If you work in a regulated role, run the publication name and About copy past your firm’s social media policy before you go live. It is far easier to clear it now than to unpick it later.

Week three: write three flagship issues before you publish one.

Do not launch with an empty pipeline. Choose three questions your ideal client actually asks you, and write a proper answer to each. A bank of three issues means you launch with momentum and buy yourself breathing room for the weeks when client work takes over. These are your pillar pieces, the ones that show how you think and that you will point people back to for months.

Week four: tell people, then publish.

Publish issue one. Send a short, personal note to twenty people who already know your work and would genuinely find it useful. A personalised message to someone who reads your posts is worth more than a blanket announcement to five hundred. Then share a short excerpt on LinkedIn with a single, low-pressure invitation to subscribe.

From week five: protect the rhythm.

One flagship issue a week is plenty. Block a recurring 60 to 90 minute slot in your calendar to draft it, and turn each issue into two to four LinkedIn posts rather than starting from scratch each time. Consistency is what compounds. Forty steady issues over a year will always beat a six-week sprint that ends in silence.

If you’re a time‑poor professional who lives on LinkedIn and is Substack‑curious, my advice is simple: treat LinkedIn as the place people discover you, and Substack as the asset that compounds over time.

You don’t need more platforms; you need one clear audience, one realistic cadence – and the courage to publish the first issue.

Subscribe now