Somewhere over the Atlantic, I stopped following the plot and started following the machine underneath it. What caught me was the scrambling. A story breaks, and within hours there is a crew on a plane. Correspondents, producers, a satellite feed, hotel rooms in a city nobody planned to be in that week. It happens constantly, and everyone on screen treats it as normal.

Someone is paying for that.

I could not work out who. I don’t have cable. Nobody I know has cable. I have never once in my life met a person who pays for a news subscription. The money has to come from somewhere, and I had no idea where.

That question took me a week to answer. The answer explains everything else about the show that confused me: the manufactured controversies, the fake partnerships, the fact that they score anchors on how sincere they seem.

We start at the screen and work backwards to the money. The jet comes last, because the answer means nothing until you’ve seen the rest.

Subscribe now

1: The Anchor

1.1 One person gets paid more than the entire investigative unit

George Stephanopoulos reportedly signed a ten-year deal with ABC in 2021 worth around $105 million. Estimates put him between $15 and $18 million a year. Michael Strahan is reported at roughly $17 million. When Craig Melvin took the Today co-anchor chair in 2025, reports put his new deal at $5 to $6 million a year.

A news producer in New York City averages about $95,000. The typical range runs from roughly $73,000 to $124,000. At Fox News in New York, the average sits around $109,000.

So the gap between the person reading the words and the person building the show that surrounds them runs somewhere between 150x and 200x.

Watch the show, and you see it everywhere. Chip is in the building before everyone and leaves after everyone. He knows every story, every guest, every problem in the rundown. He holds the whole thing up. The business also considers him replaceable, and the person sitting on the couch is not.

The structure produces that asymmetry on purpose.

1.2 The audience relationship lives on a face

Morning news is a habit product. People don’t wake up and decide to consume information. They turn on a screen while making coffee and the same person is there, the way the same person has been there for eleven years.

That habit attaches to a human being. It doesn’t attach to the network, the studio, the graphics package, or the reporting.

Which creates a specific problem. The most valuable asset in the building has legs. An anchor can walk across the street to a competitor and take a real chunk of the audience along. A producer can’t. A correspondent mostly can’t. A foreign bureau definitely can’t.

Everything strange about how these people get treated follows from that one fact. The anchor gets paid like a scarce asset because they are one. Everyone else gets paid like infrastructure.

1.3 They score the sincerity

This is the part that made me sit up on the plane.

Networks run quantified likability research on their on-air talent. Q Scores measure familiarity and appeal, and the industry uses them to decide who is worth what.

When Katie Couric’s Q Score fell to 10 among women in 2013, with 21% of them viewing her negatively, that number moved real business decisions.

Sit with what that means. There is a number attached to how much you seem like a person worth trusting. It gets tracked. It rises and falls. Your contract follows it.

The job is to be yourself at a level that tests well.

1.4 The research has been steering coverage for a very long time

Magid is a media consultancy that has advised TV news for decades. They had a hand in building the format that became Good Morning America.

In 2003, Magid released a survey to its clients reporting that of all topics tested among 6,400 viewers, war protests came in dead last.

The United States was at war at the time. The research said the protests against it were the least interesting thing on the menu.

Nobody had to issue an order. The number was the order. Cover what tests poorly and you take a commercial risk, and the person taking that risk has a boss.

Hold onto Magid. They come back at the end of this piece with something much more direct.

Leave a comment

2: The Bookers

2.1 The invisible supply chain

The show doesn’t sell information. You can get information anywhere, instantly, free.

It sells exclusivity. Getting the person at the centre of the story onto your couch instead of the couch two blocks away.

The people who do that are called bookers. Formally, they’re segment producers. The Washington Post described the job as persuading a suddenly famous nobody to tell their story on air to millions of strangers, ideally from a studio in New York.

Speed decides it. Being first on the scene often means locking an exclusive that denies your competitor the same interview. Jonathan Wald, who has run Piers Morgan Tonight and the Today show and sat second in command at CNBC, described bookers to NPR as “snipers who sit waiting for their prey.”

The Post also noted, plainly, that bookers often pay for those heads.

2.2 The trade nobody says out loud

The guest has something the show needs. The show has something the guest needs. Sitting between them is a publicist whose entire job is protecting their client’s image.

Publicists routinely ask in advance what questions will be asked, to prevent surprises and keep unwanted topics off the table.

That’s the deal. Access in exchange for framing. Nobody signs anything. Everyone understands it.

Now run it forward. Burn a publicist with a hard interview, and that publicist has fourteen other clients and three other networks to hand them to. A genuinely tough segment doesn’t cost you in that segment. It costs you across the next two years of bookings.

So the manufactured partnerships you see on screen come straight from life. They’re the visible surface of a procurement relationship.

2.3 The get is worth one morning

A booking moves the number for a single day. Maybe two. Then it’s gone, and the machine needs another one tomorrow, and the day after, and every day, forever.

That treadmill explains why bookers behave the way they do. It also explains why the show manufactures conflict internally. Real news doesn’t arrive on a schedule that fills four hours a day, five days a week, fifty-two weeks a year. Something has to.

3: The Show's Own Books

3.1 Morning television is a printing press

Morning shows sit on some of the most profitable real estate in broadcast.

The reasons are boring and powerful. Four hours of advertising inventory a day. Cheap to produce compared to almost anything else on television. And an older, wealthier, more attentive audience that advertisers pay real money to reach.

The show you’re watching funds the expensive parts of the news division.

3.2 The lifestyle segment pays for the war coverage

Kagan’s estimates give a sense of scale on what news operations spend. In 2022, Fox News put roughly $1.48 billion into newsroom expenditures. CNN spent about $882 million. MSNBC spent about $448 million.

Foreign bureaus, investigative units, evening broadcasts, and breaking news coverage don’t pay for themselves. They never have.

The cooking demo at 8:40 is the reason the news exists.

Which explains where the cost-cutting lands. GMA shaved about $100 million off the balance sheet by giving up its Times Square studio. Hoda Kotb reportedly left Today after being told her pay would be cut. Anderson Cooper settled for less than he was making.

In February 2026, the Washington Post laid off roughly a third of the company. Sports, books, podcasts, foreign desks, business, national.

3.3 The question this raises

If a morning show is that profitable, and it funds a news division that isn’t, the whole structure rests on one thing staying true.

The morning show has to keep making money.

So where does that money come from? Because it isn’t from you, and it isn’t from me.

Share

4: The Network

4.1 The best subsidy machine ever built

This took me the longest to understand, and it’s the most important idea in the piece.

For about thirty years, people who never watched cable news funded almost all of it.

The cable bundle worked like this. You paid your provider somewhere around $90 to $100 a month. That money got divided among a hundred-odd channels through negotiated deals called carriage fees. Every channel took its cut based on what it could extract in the room.

Fox News collected somewhere between $1.72 and $2.18 per subscriber per month. Every subscriber. Around 90 million households.

Roughly 14% of cable subscribers watched Fox News with any regularity. The other 86% paid anyway. Techdirt ran the arithmetic and put it starkly: about 87 million Americans were funding the channel while rarely or never watching it.

Think about what that does to a business. Your revenue has almost nothing to do with whether anyone watches. Ratings matter for the advertising layer sitting on top. The floor underneath is guaranteed by strangers who don’t know they’re paying you.

4.2 The natural experiment

If you want proof that the subsidy is the business, compare four channels in the same genre in the same year.

Kagan’s 2022 estimates:

Fox News. Revenue $3.26 billion. Profit $1.78 billion.

CNN. Revenue $1.77 billion. Profit $892 million.

MSNBC. Revenue $903 million. Profit $455 million.

Newsmax. Revenue $66 million.

Newsmax had zero license fee revenue. Virtually all of its money came from advertising.

Same industry. Same year. Same product category. Three channels inside the bundle, one outside it. The one outside it made two per cent of what the biggest one inside it made.

That’s a plumbing gap.

4.3 Loyalty converts directly into pricing power

There’s a second-order effect here that’s a real business lesson.

Fox News collected a higher per-subscriber fee than CNN and MSNBC because of what it could threaten in a negotiating room.

Drop Fox News, and you face subscribers who cancel their entire cable package over it. Drop CNN, and you face subscribers who mostly shrug.

Fox knew this and priced accordingly. They pushed for $3 per subscriber, which at scale would have meant over $1.8 billion a year from carriage fees alone.

Audience intensity shows up as a number in a contract the audience never sees.

5: The Parent Company

5.1 The news division is a rounding error

Every news operation you can name sits inside something much larger that mostly cares about other things.

The people who own news frequently don’t care about news. They care about a portfolio. News is a prestige asset, a political asset, a regulatory chip, and a line item.

Which means decisions get made about the newsroom in rooms the newsroom is not in.

5.2 What happened to MSNBC

You don’t need a hypothetical for this. Look at the last eighteen months.

In January 2026, Comcast completed the spinoff of most of its cable networks into a new public company called Versant. That company holds CNBC, USA Network, E!, Syfy, Golf Channel, Oxygen, and MSNBC. Roughly $7 billion in annual revenue.

As part of the split, MSNBC was renamed MS NOW and stripped of the NBC peacock. Versant’s CEO explained in a staff memo that the peacock is synonymous with NBCUniversal, and NBCUniversal decided to keep it.

A thirty-year-old news brand lost its name and its logo because of a corporate asset allocation. NBC News cut about 150 jobs, roughly 2% of the division, ahead of the separation.

Then in June 2026, Comcast announced it would spin off NBCUniversal and Sky as well. Around the same time, Paramount Skydance won Department of Justice approval for a $110 billion deal for Warner Bros. Discovery, which owns CNN.

Nobody asked the audience about any of it. The audience isn’t a party to these transactions. Neither is the newsroom.

The corporate anxiety running through The Morning Show is the most accurate thing in the show.

Subscribe now

6: What Broke

6.1 Cord cutting, explained properly

Everyone tells this story wrong. The usual version says audiences left for streaming.

Watch who left first.

The subsidy came from people who paid for cable and never watched news. Those are exactly the people with the least reason to keep paying for cable. They were handing over $100 a month for a hundred channels to watch four of them.

So the non-viewers cut the cord first. The money walked out before the audience did.

US pay-TV penetration peaked around 82 to 83% in 2011. Today it’s roughly 34% of households. In 2011, a cable advertisement reached a majority of American homes. Now it reaches about a third.

Comcast alone shed 1.15 million TV subscribers in 2025. That’s roughly 3,500 people a day, every day, for a year.

The audience for cable news is still there. It’s older, and it’s loyal, and it still shows up. The eighty-odd million strangers quietly funding the operation are the ones who left.

6.2 CNN+ is the cleanest autopsy in modern media

When the bundle started failing, the obvious answer was to sell news to the people who actually wanted it. Cut out the subsidy, charge the fans.

CNN tried it. The result is one of the most instructive failures I have ever read about.

CNN+ launched on March 29, 2022. About $300 million spent, with a plan to invest $1 billion over four years. Executives projected 2 million subscribers in year one and 15 to 18 million after four.

Two weeks after launch, fewer than 10,000 people were using it daily.

It shut down on April 30. Thirty-two days. It had around 150,000 subscribers when it died.

They missed their first-year projection by a factor of thirteen, and they found out in a month.

The lesson is brutal and specific. News is something people bump into. For decades the bundle guaranteed the bumping-into, and everyone mistook that for demand.

Take away the accident and most of the demand was never there.

6.3 The replacement channel is collapsing too

After print died and cable started sliding, publishers rebuilt on search traffic. Google sent readers, readers saw ads, the machine turned over.

That’s going the same way.

Chartbeat data shows Google organic traffic to publishers down 33% globally and 38% in the US between November 2024 and November 2025.

A Growtika analysis tracked ten major tech and media outlets from early 2024 to January 2026. Their combined US Google traffic fell from 112 million monthly visits to just under 50 million. Digital Trends went from 8.5 million clicks a month to 264,861. A 97% collapse.

Between June 2025 and June 2026, USA Today’s national site lost close to half its Google traffic. Business Insider lost more than 85%. CNN around 25%.

Roughly 60% of all searches now end without a click to any website. On mobile it’s 77%.

And the thing everyone assumed would replace it hasn’t shown up. Referrals from ChatGPT and similar tools currently make up under 1% of publisher page views.

6.4 So nobody is paying, and nobody trusts it

The Reuters Institute’s 2026 Digital News Report surveyed nearly 100,000 people across 48 markets. The findings answer my question on the plane directly.

Across 20 tracked markets, 17% of people pay for online news. That’s flat against 18% the previous year, and the US figure fell 4 points.

You have never met someone who pays for news because roughly five out of six people don’t.

Trust in news in the US sits at 25%, the seventh lowest of 48 countries surveyed. Among right-leaning Americans, it’s 15%.

News avoidance globally is at 42%, up from 29% in 2017. The share of people describing themselves as very interested in news fell from 59% in 2021 to just under 46%.

And for the first time, social media and video networks passed publishers’ own sites as a way to access news. 54% against 51%. Among 18-to 24-year-olds, more than half say social, video, or AI is their main news source.

6.5 Magid comes back

Remember the consultancy from section 1.4, the one that told clients in 2003 that war protests tested dead last.

In April 2026, Magid’s chief operating officer Jaime Spencer told Variety what drives the business now.

The attributes that work: insightful, thoughtful, reassuring, timely, calming, innovative.

The attributes losing relevance: balanced, explanatory, trustworthy, essential, substantive, reliable, confident, accurate, clear.

He identified that second list himself as tenets of journalism, and said brands built around them are unlikely to break through.

Then he said this: if you are looking to monetise and grow an audience, trustworthiness isn’t really part of that anymore. Content that confirms a viewer’s existing beliefs gets more attention, more consumption, and stronger brand connection.

He also told them there is no correlation between overhead costs and perceived quality, and that local news operations should stop developing a bench of stars and bet on one main quarterback instead.

Read that back against the 2003 survey. Twenty-three years, same firm, same logic, better vocabulary.

Nobody is hiding this. It’s in a trade publication, on the record, from the consultancy that helped invent the format.

Share

7: So Who Actually Pays for the Jet

Three sources. None of them is a person buying news.

7.1 Retransmission fees

Cable and satellite providers pay local broadcast stations for the right to carry their signal. Those payments are enormous, and almost nobody outside the industry knows they exist.

Nexstar’s 2024 revenue was $5.4 billion. Of that, $2.9 billion came from retransmission. Fifty-four per cent of the business.

Across broadcast station groups generally, somewhere between a third and half of annual revenue comes from these fees. About 65 million US households still buy some kind of linear TV bundle, and every one of them is feeding this.

This is why Sinclair and Nexstar spent a decade buying up local stations. Boring, unglamorous, enormously cash generative.

7.2 Politics

Every two years, a wave of money arrives that has nothing to do with viewers or advertisers in any normal sense.

AdImpact projects $11.6 billion in political advertising for the 2025-26 cycle, with roughly $5.6 billion of it going to broadcast television. S&P Global puts local TV political ad revenue at $4.02 billion in 2026, which would be 16.3% of net total broadcast revenue. A record for a non-presidential year.

Campaign spending substantially funds local news in America, on a two-year cycle, concentrated in whichever states happen to be competitive.

7.3 The bundle, rebuilt

The New York Times figured out something the rest of the industry is still catching up to.

In Q2 2026, the Times added 280,000 digital subscribers to reach 13.4 million. Digital subscription revenue rose 16.4% to $408 million. Digital advertising was up nearly 21%. They crossed $2 billion in annual digital revenue in 2025, and they’re targeting 15 million subscribers.

Look at what people are buying. Wordle. The crossword. Cooking. The Athletic. Then the news, bundled in with all of it, at a price that stepped up from $25 to $30 for long-tenured subscribers with retention holding.

The Times rebuilt the cable bundle inside one company. Games and recipes hold the door open, and the reporting lives inside.

That’s structurally the same trick a morning show runs. Cheap, pleasant, high-margin content carries expensive, unprofitable, important content.

7.4 And the part that isn't funded

Meanwhile: layoffs across US and UK journalism topped 2,300 newsroom jobs in the first half of 2026 alone, running about 34% ahead of the previous year’s pace. CBS News shut down its century-old radio division.

Roughly 2,900 newspapers have closed in the United States since 2005. More than 200 counties now have no local news outlet at all.

So here’s the honest answer to my question. The jet gets paid for by a subsidy from people who never watch, plus a fee attached to a bundle most people have already left, plus campaign money that arrives every second year. And in the places where none of those three things reaches, the jet stays on the ground.

8: What This Actually Teaches

Five things. Every one of them comes out of something above.

The product is manufactured sincerity, and it gets scored. What’s being sold is the feeling that a person on screen is real. It gets measured, priced, negotiated by agents, and optimized by consultants. The one product that can’t survive being manufactured is being manufactured at industrial scale.

Value creation and value capture came apart, and the gap is 200x. Chip builds the show. The anchor owns the audience relationship. Only one of those is portable, and portability is what gets paid. If your contribution can’t walk out the door and take customers with it, you get priced as infrastructure no matter how essential you are.

A subsidy hides which parts of your business anyone wanted. For thirty years, cable news never found out. Newsmax made $66 million doing roughly what Fox did for $3.26 billion. The difference was plumbing. When your revenue is disconnected from whether anyone shows up, you’re running blind, and you won’t know it until the plumbing changes.

Loyalty is a number in a room you never enter. Fox charged more per subscriber than CNN because Fox viewers would cancel their cable over it and CNN viewers wouldn’t. Audience intensity converted straight into pricing power in a negotiation the audience never saw.

The optimization consumed the thing being optimized. An industry spent decades tuning itself toward feeling trustworthy. Trust in US news is now 25%, and 15% among people on the political right. Seventeen per cent of people will pay for it. The consultants have responded by advising clients to stop leading with accuracy and reliability, because those attributes no longer move the business.

It’s 1:14 AM. You’re adjusting the kerning on a slide nobody will read past slide two.

Three feet away, your phone is lit up. A message from a customer worth $40,000 a year, sent six hours ago. Still unread.

You’ll get to it after this slide.

You said that last night too.

That’s not a discipline problem.

Every founder says the same thing about delegation. “I know I should do it more.” Then they don’t.

Nobody fails at delegation because they lack willpower. They fail at it because they’ve never actually priced their own time the right way, and because letting go of a task feels like losing a piece of who they are or losing control of their organisation.

Fix the first problem and the math forces your hand. Fix the second, and you stop needing willpower at all.

The Real Cost of Not Delegating

Start with the math, because it’s the part everyone skips.

If you have not read last week’s newsletter, I would recommend you read that first. If you have, you already know the trick: your hourly rate isn’t what your invoice says. It’s what your invoice says, adjusted for the odds you actually get paid at all. Early-stage founders especially are taking on a massive risk premium just by staying in business another year, which means the true rate on your time is a multiple of anything you’d quote a client.

Here’s what that means in practice. If your real hourly rate is $500, and you’re spending three hours a week formatting invoices or booking your own travel, it might feel like you are saving some money by doing it yourself, but in reality, you’re paying yourself $500-an-hour to do a $20-an-hour job.

Nobody would take that trade if you wrote it out like that. Most founders take it every week without noticing, because the cost never shows up as an invoice.

The Actual Blocker Isn’t Logic

Ask a founder why they still do their own scheduling, and you’ll get a logical-sounding answer: “It’s faster if I just do it myself.” That’s just not true, and the founders know it. They say it anyway because the real reason is harder to admit.

Delegating means admitting someone else can do a piece of your job. If your identity is built around being the person who personally solves every problem and who works the hardest in the building, handing off a task doesn’t feel like freeing up three hours. It feels like losing a piece of that identity.

This is worth naming honestly, because it’s the actual thing in the way.

Another thing, which, according to me, is the more important thing, is trust. Founders do not trust anyone else to do their job better than them. After all, they are the founder. But even if you take it to be true, the job you are delegating is not worth your hourly rate. Trust is built through the ability to do the task well. Anyone you hire is realistically going to get so good at what they do in a couple of months because of repetition that you will start trusting them more than yourself to do the same job. That’s the same way with people who get executive assistants. At first they can’t imagine letting these tasks go, and after that they can never go back to doing the tasks themselves.

Name the actual blocker first. Then the frameworks below actually work. Skip it, and you’ll read the frameworks, even agree with them, but change nothing.

Finding What to Delegate

Most founders get delegation wrong. They hand off whatever annoyed them most this week, which is fine too, but that means they’re reacting instead of deciding. Here is the framework I would use instead:

Step 1: Audit a real week. Write down everything you did, hour by hour. I do this myself from time to time to figure out where my time is going. The best way I have found is to use your current calendar app. Set an alarm after every hour and write down in your calendar itself what you did for that hour.

Step 2: Sort into three buckets.

• Founder-only: Stuff that you would not be willing to delegate because this drives real results and moves the needle for your business. Vision, key hires, fundraising, the handful of customer conversations that actually need you to be in the meeting.

• Trainable: needs judgment, but a capable hire can learn it by shadowing you and learning how you think. More tasks fall here than you would think.

• Transferable: recurring, process-driven, no judgment call required.

Step 3: Run this four-question test on anything left. A single “no” is enough to delegate it.

• Is this one of the few things that meaningfully moves the business forward, and are you specifically positioned to do it because of your judgment, relationships, or role? (Peter Drucker’s contribution test, from The Effective Executive.)

• Are you operating at your best skill here, the kind of work you’d put in your own top tier, not just competent, but excellent? (Dan Sullivan’s Unique Ability framework, from Who Not How.)

• If someone else did this with clear instructions, would the result measurably suffer? (Tim Ferriss’s rule, from The 4-Hour Workweek.)

One “no” and it’s an immediate delegation.

Step 4: Automate or eliminate before you delegate. Don’t build a handoff process for something you could kill outright, or a Claude Cowork task could replace entirely.

Who Actually Does the Work

Once you know what to hand off, most founders default straight to “hire someone.” That’s usually the wrong first move.

The real spectrum is wider than it used to be:

• Contractor or freelancer — for defined, scoped work with a clear deliverable.

• Virtual assistant — for recurring admin: scheduling, inbox triage, travel, expense tracking.

• Fractional hire — for judgment-heavy work you need part-time, not full-time. Fractional CFOs are really successful right now.

• Full-time hire — only once a role has enough recurring, defined work to justify managing a person.

• AI agent — for anything repetitive, rules-based, or synthesis-heavy: research, first drafts, data pulls, scheduling logistics.

That last one is the newest lever, and the one most founders are underusing or misusing. It is very exciting for founders to get on AI agents either by building them themselves or getting someone to build them for them. Because of that excitement, they trust the AI agents’ abilities more than they should. Handing an agent your calendar logistics is low-risk, but handing it the authority to negotiate a vendor contract or respond to an upset customer without review is how a small error turns into an expensive one. For those kinds of tasks, I would still recommend one of the other methods.

The rule that holds up either way: delegate mechanics first, judgment much later. A new hire doesn’t get to make pricing calls in week one. Same logic applies to an agent.

The Part Everyone Skips: How to Hand It Off

This is where most delegation attempts die, and it has nothing to do with who you picked.

Founders hand off a task with a verbal explanation and a vague sense of the outcome, then get frustrated two weeks later when the result is wrong. The problem was never the person. It was that nobody wrote down what “good” actually looks like.

Documentation beats trust, every time. Not because trust doesn’t matter, but because you can’t build it off a Slack message. You build it off a process someone can run without pinging you every hour.

Before you hand anything off:

• Write down the outcome you actually want, specifically. Not just something super vague like “handle the newsletter,” but tell them exactly what subject line style you prefer, what length and even what gets cut.

• Sit through one round of questions before the handoff. Make sure the person you are handing off to leaves the room with no further questions.

• Schedule the first review meeting after the first round of work has been completed. The quicker the feedback loop, the better they get.

• Build in a buffer between the review and the real deadline. Plan for the redo. Be pleasantly surprised when you don’t need it.

Where Founders Get This Wrong

A few patterns show up over and over, at every stage:

• Delegating without documentation, then blaming the hire, or the agent, when the result misses.

• Giving too much authority too early, especially to AI agents, because the first few results looked good.

• No review loop, so small errors compound silently until they’re expensive.

• Confusing being busy with being right to delegate. If you’re working 60-hour weeks and still the bottleneck on everything, the answer isn’t more grit. It’s picking one specific thing to hand off this week. Use the exact framework we discussed to hand it off.

There’s a real tension worth naming instead of ignoring: the math says delegate aggressively, but early-stage founders are usually cash-constrained and not time-constrained by choice. That’s exactly why the cheap options are better than the expensive ones early on. An AI agent handling research and first drafts costs a few hundred dollars a month. The same function as a full-time hire costs tens of thousands a year. So the AI agent is the way to go until you can actually get to the full-time hire.

The One Thing to Do This Week

Pick one task from your audit, the one you dread most and the one that failed all three questions in the filter above. Write down exactly what “good” looks like. Hand it to a person or an agent and schedule the review.

You don’t have a time management problem. You have a pricing problem. Now you know the price.

Try this: picture your business posting the job you actually do, as if it were a normal job listing for an outside candidate.

No sane candidate applies to that listing. And yet this is the exact job every founder already has, most without ever writing the terms down or asking whether they’d take it if a stranger offered it to them.

Here’s the part that should bother you more than it probably does: everyone else in the building is protected from this. Your employees get paid on time, every two weeks, regardless of whether the company is having a good quarter or a terrible one. If the business folds tomorrow, they update their resume and take the next job. You do not get that option. You are first in line to lose money and last in line to get paid. You are structurally the highest-risk person in your own company, and most founders still price their own hour like it’s the safest job in the room.

So let’s work out what that hour should actually be worth. And then, because the honest number is going to be uncomfortably large, let’s work out the much more useful question: below what number should you actually be handing work off, because obviously you cannot delegate every task priced under your real rate. Nobody runs a company that way.

What would you have to pay someone else to take your seat?

Suppose the exact same skillset, at the exact same skill level, is worth $50/hour as a stable job somewhere else. Reliable paycheck, every two weeks, guaranteed.

Now ask: if you had to hire a rational person to sit in your seat instead, on the actual terms you’re working under, meaning they don’t get paid until the company can afford it, and there’s a real chance the company never gets there, what would you have to offer them to take the job?

Nobody rational takes that trade at $50/hour. The gap between $50/hour and whatever number would make them take it is the real price of the risk you are personally absorbing every single day you run this thing.

The boring adjustment first

Part of the gap is just the time value of money. A dollar four years from now is worth less than a dollar today, since you could have invested today’s dollar instead. Even at generous interest rates, this only nudges $50/hour up to something like $53 to $55. Real, but not the story.

The adjustment that actually matters: you might get paid nothing

The real problem isn’t that the payout is late. It’s that the payout might never come.

Suppose there’s a 25% chance the business survives long enough to pay this seat what it’s actually owed. To make the expected value of taking this seat equal to the safe $50/hour job, the math is:

In plain English: if you only get paid once out of every four times you take a seat like this, that one payout has to be four times bigger, just so that, averaged across every time you’d take this bet, you land back where you started.

That’s the break-even rate. Most people stop here and assume $200/hour is the answer. It isn’t, and this is where it gets interesting.

Why break-even still underpays you

Picture four different people, each taking this exact seat, each priced at the $200/hour break-even rate. Statistically, three of them get nothing after years of work. One gets paid out at four times the rate.

Add up the outcomes across all four, and the total money paid out is identical to what would have happened if all four had simply kept their safe $50/hour jobs the whole time. Collectively, nobody is better off for having taken the risk.That’s the catch with break-even pricing. It’s fair the way an actuarial table is fair, averaged across thousands of repetitions. But you don’t get thousands of repetitions of your own company. You get one. You don’t get to be “the average of four founders.” You get to be exactly one of them, and you find out which one only at the end.

A rational person does not accept a bet that merely breaks even, with a real chance of walking away with zero, when a safe alternative sits right there. To make the risky seat worth choosing over the safe job, you need to charge more than break-even. This is the same logic behind why stocks have to return more than a savings account over time (the equity risk premium), and why a venture fund needs a small number of enormous wins to justify writing off most of its bets. The premium above break-even is the actual price of the uncertainty, not the price of the hours worked.

Now stretch it out year by year, because that’s where it gets brutal

The 25% figure above was just a starting assumption. Real survival odds compound, year over year, and this is where most founders badly underestimate how fast the risk stacks up.

Suppose your business has a 20% chance of failing in any given year. That’s an 80% chance of making it through each year, which sounds almost comfortable. One-in-five odds of dying in a given year feels manageable.

But survival compounds multiplicatively, not additively. Needing to survive several years in a row means multiplying that 80% by itself, once per year:

An 80% annual survival rate, which sounds almost safe, still turns into roughly a 1-in-3 chance of surviving five straight years. The break-even rate more than triples over that same stretch, and it isn’t a straight line. It curves, because each additional year doesn’t just add risk on top of the old risk, it multiplies against everything that came before it.

Putting it together: what your hour is really worth

Take the four-year break-even figure: about $122/hour, just to match the safe $50/hour job in expectation. Now layer the risk premium argument from before on top of it, since break-even alone still isn’t real compensation. A defensible target for the seat you actually occupy, given four years of real survival risk, lands somewhere around $240 to $250/hour.

You obviously cannot delegate everything under $250/hour

Here’s where the math, taken at face value, breaks down in practice. If your real rate is $250/hour, the naive conclusion is “delegate anything that costs less than $250/hour to do.” Followed literally, that means hiring someone for nearly every task in the business, since almost nothing you’d outsource costs anywhere near $250/hour. That’s not leverage. That’s just spending money you probably don’t have, on the theory that your time is precious.

So the real question isn’t “what is my time worth.” It’s: below what number does delegating a task actually make sense once you account for the fact that freed-up time doesn’t convert cleanly into more $250/hour work?

Here’s the honest math behind that.

When you hand off a task, you don’t get to bank the full $250/hour value of the hour you freed up. Three things eat into it:

1. You don’t have infinite $250/hour work waiting. Most founders, especially early on, don’t have a bottomless queue of top-value work sitting ready the instant an hour opens up. Realistically, maybe half of a freed hour actually gets reinvested into something at your top rate. The rest goes to lower-value catch-up, planning, or just doesn’t get captured at all. Call this the capture rate, roughly 50%.

2. Delegation isn’t free of your time either. Explaining the task, checking the output, answering questions: call this the oversight tax, typically 15 to 20% of the time you thought you were saving.

3. Someone new gets things wrong sometimes. Rework, corrections, and the occasional redo eat further into the gain: call this the quality tax, another 10 to 15%.

Multiply the frictions together:

Rounded down conservatively, once you account for the fact that early on these frictions run worse than best-case, you land close to 25%. So the realistic delegation threshold isn’t your full rate, it’s roughly a quarter of it:

This isn’t just a made-up shortcut. It’s the same conclusion Dan Martell arrives at in Buy Back Your Time, where he calls this number the Buyback Rate: take your effective hourly rate and divide it by four, and that’s the ceiling below which outsourcing a task is close to a no-brainer, even before you get precise about frequency and quality.

The difference here is what you divide. Martell’s version starts from your nominal pay, income divided by hours worked, with no adjustment for the risk you’re actually carrying. Our version starts from the risk-adjusted rate, the $250/hour figure that accounts for the real odds you never get paid at all. That changes the answer meaningfully:

Which, interestingly, lands almost exactly back where this whole exercise started: the same neighborhood as the $50/hour safe job you were comparing yourself against in the first place.

That’s not a coincidence, and it’s the most useful takeaway in this whole piece: most founders anchor their delegation threshold to what they currently pay themselves, which is usually far below what the risk they’re carrying actually justifies. Price the risk correctly first, and the honest delegation threshold comes out noticeably higher than most founders assume, which means most founders should be delegating more aggressively than they currently are, not less.

The math tells you the number. It doesn’t tell you how to actually find the right task, the right person, and the right process to hand it to without creating a new mess to clean up later. That’s next.