A few weeks ago I was on a call with a PM I coach. Sharp. Eight years of experience. Real shipped products, real outcomes. She’d just come out of an onsite loop for an AI PM role, and the first thing she said was:

“Nazuk, there was a round I’d never even heard of. They put me in front of an AI coding environment and asked me to build something. I’m a PM. Nobody told me this was part of the interview now.”

She’d prepped hard. She’d done the mock product sense questions, rehearsed her STAR stories, memorized her metrics frameworks. She walked in confident and got blindsided anyway. Not because she was a weak candidate, but because she prepped for an interview that no longer exists.

I’ve been in tech product management for almost 15 years, and I’ve spent the last five coaching aspiring and mid-career PMs. I have never seen the interview change this fast.

Not in the 2015 “design a product for the blind” era, not in the 2021 hiring frenzy. The change happened quietly and companies continue to evolve their hiring processes, over the last eighteen months, and most candidates found out the hard way, mid-loop.

So consider this the memo nobody sent you. We’re going to walk through what actually changed (straight from the companies themselves, not interview-prep folklore), why your old prep quietly stopped working, and then go round by round through the modern AI PM interview stack: all seven rounds, what each one is really testing, and one concrete prep move for each.

It’s a long one. That’s deliberate. Save it, forward it to the friend, partner who’s been applying into the void, and let’s go.

Section 1: What changed?

When I say “the interview changed,” I don’t mean interviewers got moodier. I mean companies rewrote their hiring processes, publicly, in writing.

Shopify made AI fluency a hiring gate. In April 2025, CEO Tobi Lütke sent a memo, which he then posted publicly, declaring that “reflexive AI usage is now a baseline expectation at Shopify.” Teams can’t even request a new hire without first proving AI can’t do the job, and AI proficiency is now part of performance reviews. His exact words: “I don’t think it’s feasible to opt out of learning the skill of applying AI in your craft.” Read that again as a job seeker. If AI usage is a baseline expectation for employees, it’s being screened for in candidates, whether the job description says so or not.

Canva doesn’t just allow AI in interviews; it requires it. In June 2025 Canva’s engineering leadership published a blog post with a title that tells you everything: “Yes, you can use AI in our interviews. In fact, we insist.” Candidates are expected to bring tools like Copilot, Cursor, or Claude into the technical interview. And here’s the detail most people miss: because AI makes the easy parts easy, Canva redesigned the questions to be more complex, ambiguous, and realistic. The bar didn’t drop. It moved.

Meta rebuilt one of its coding rounds around AI. Starting in late 2025, Meta rolled out an AI-enabled coding interview where candidates work alongside an AI assistant in the interview environment, and they get evaluated on problem solving, code quality, verification, and communication. Notice “verification”: they’re explicitly scoring whether you can check AI’s work, not just accept it.

Google is piloting the same thing: an AI-assisted coding interview with an approved assistant available to candidates.

Anthropic wrote candidates an actual instruction manual. Anthropic initially banned AI in applications, then reversed course and published official guidance on candidates’ AI usage: use AI freely to refine your resume and prepare for interviews, but expect live interviews to be AI-free unless told otherwise, because they want to see how you think in real time. Sit with that combination for a second, because it’s the whole new world in one policy: use AI to prepare, then demonstrate you can think without it. Both skills. Screened separately.

And while the interview was changing, the market moved too. AI product roles have nearly quadrupled globally since 2022, with over 12,000 AI product jobs posted in the US in the first half of 2026 alone. But look inside those numbers: 47% of AI product postings are manager-level and only about 2% are junior. Companies are hiring people who can own AI product direction, which means the interviews are designed to find out, quickly and sometimes brutally, whether you actually can.

So that’s the situation. More AI PM jobs than ever, interviews explicitly redesigned for the AI era, and almost nobody prepping for the interview as it now exists.

Section 2: Why your old prep stopped working

Here’s the uncomfortable mechanism underneath all of this.

The old PM interview meta was, frankly, gameable. There were known question banks. There were frameworks: CIRCLES, AARM, name your acronym. You could rehearse forty answers, pattern-match the question in the room, and deliver. Interviewers knew this, but it mostly worked as a filter anyway.

Then two things broke it at the same time.

First, AI made polished generic answers free. Every candidate now has access to the same tools that generate the same competent-sounding resume bullets and the same framework-shaped answers. Interviewers read hundreds of AI-smoothed applications that all sound alike. The polish that used to differentiate you is now the baseline, which means the differentiator moved to the one thing AI can’t fake for you: specific evidence from your actual career, delivered with real-time judgment.

Second, companies started testing what the job actually became. The AI PM job in 2026 involves prototyping with AI tools, scoping AI features with genuinely uncertain behavior, and making judgment calls about what to ship. So interviews evolved to sample that work directly, which is where the new rounds come from.

Section 3: 6-7 rounds of the AI PM interview

Not every company runs all seven, and the order varies. Some small-to-mid-size companies may only have about 3-4 rounds of interviews.

But this is the full stack I now see across my coaching clients’ loops, and you should know what each round is actually scoring before you walk into any of them.

Round 1: The recruiter screen

What it looks like: A 20–30 minute call that feels like a friendly chat but is not.

What it’s really testing: It’s not a chat. It’s a filter, and it’s usually decided in the first five minutes. Recruiters are typically not PMs; they’re matching what you say against a checklist, under time pressure, with a stack of other candidates behind you. Your job isn’t to impress them. It’s to make their job easy: can they place you instantly (”PM with 8 years in B2B SaaS, shipped AI features into enterprise workflows”), does your story use the language of the JD, and do your resume and LinkedIn tell the same story when they cross-check?

Write your one-line intro as role + domain + proof and test it on someone who isn’t a PM. If they can’t repeat back what you do after hearing it once, it’s not done.

Round 2: The hiring manager round + Behavioral

What it looks like: A conversation with the person whose team you’d join, walking through your background.

What it’s really testing: Whether your career experience maps to their problems. A hiring manager listening to your background is running one loop in their head: “does this person solve what I need solved this year?” Most candidates recite their history chronologically. Strong candidates do the mapping for the manager. They’ve researched the team’s product, challenges and surface the two or three pieces of their experience that line up with it and lead with that answer.

Before you interview with the hiring manager, write down your best guess at the manager’s top three problems (the product, the JD, and their LinkedIn activity will get you 80% there). Then choose which of your stories answers each. Walk in with the mapping done.

These interviews incude questions like “Tell me about a time when…” a flavor of 6-7 of these questions.

What it’s really testing: Not whether you can produce a STAR-shaped answer; everyone can, and AI has made STAR-shaped answers infinitely reproducible. What *good* interviewers are actually probing is the psychology underneath: how you operate under conflict, what you do when you’re wrong, whether you take real ownership or narrate your team’s work in first person.

A good interviewer will ask you 1-3 follow-up questions untill they fully get a sense of the project/ situation you just described. Rehearsed answers survive the first question and die on the second, because the candidate memorized an answer instead of owning a story.

Pro tip: Stop building answers per question. Inventory the 6–10 defining moments of your career (the launch that went sideways, the conflict you resolved, the call you made with incomplete data) and know each one cold, including the messy parts. Six deep stories beat sixty rehearsed answers, because you can’t be knocked off a story you actually lived.

Subscribe now

Round 3: Product sense

What it looks like: “How would you improve X?” “Design a product for Y.”

What it’s really testing: Judgment, not framework recall. This is the round AI broke most visibly. Interviewers have now heard the same CIRCLES-shaped answer hundreds of times, and they discount it on sight. What scores well in 2026 is structured thinking that sounds like yours: a point of view on the user, a willingness to make a call and defend a trade-off, an instinct for what matters versus what’s noise.

For AI products specifically, expect a twist the frameworks never covered: reasoning about probabilistic features. What should the product do when the model is wrong? How do you design for trust?

I personally like this framework, where including AI or building with AI is not after the fact, but part of the whole answer along the way. You may build your solution with AI or even without it, important part is how you land at the answer in 20 minutes.

If you want, to remember a single word for this framework is discernment, which i like to call a skill PMs need to build, which is holistic thinking, not just judgement but holistic thinking.

Short explanation below

Share

Pro-tip: Practice answering out loud and record yourself; the gap between how structured you feel and how structured you sound is the whole game here. Then, for AI roles, take one AI product you use and articulate three deliberate design decisions its team made to handle model uncertainty. That’s the muscle this round tests.

Round 4: Execution + Analytical

What it looks like: Metrics questions, prioritization scenarios, “Your launch just regressed a key metric. Walk me through it.” It also could be a case study-style interview or questions probing your past experience

What it’s really testing: There’s a rubric, and most candidates have never seen it. Execution rounds score whether you show operator instincts: do you ask what the goal metric is before optimizing, do you think in trade-offs and counter-metrics, do you distinguish signal from noise, can you sequence work under constraint. Candidates fail this round not from lack of intelligence but from answering like an analyst (here are five things we could look at) instead of an owner (here’s the first thing I’d check and why).

For your current or most recent product, be able to draw the metric tree from business goal down to driver metrics, from memory. Nearly every execution question is that tree, worn as a disguise.

Round 5: Analytical + Technical

What it looks like: Data interpretation, sizing, experiment design. Sometimes a take-home, more about your technical and analytical depth

What it’s really testing: Whether you can reason through data out loud: state assumptions, do rough math without panicking, notice when a number smells wrong, and say “here’s what I’d want to check before trusting this.” Note what Meta scores in its AI-assisted rounds: verification is an explicit criterion. That’s spreading everywhere, because in an AI-saturated workflow the PM’s analytical job is increasingly to be the person who catches the confident-but-wrong output.

One prep move: Rebuild your comfort with thinking aloud through numbers. Take any dashboard you have access to, pick one weird movement, and narrate your investigation out loud for five minutes. It feels ridiculous. It works.

Round 6: Vibe coding, the round nobody told you about

What it looks like: You, an AI coding tool, and a prompt like “build a working prototype of this feature idea.” Yes, for PMs.

If you go in the interview and jump to prototype, you may not survive the round.

What it’s really testing: This is the round that blindsided my client, and it’s the clearest expression of where the industry moved. With Canva insisting candidates use AI tools in interviews, Meta building an assistant into the interview environment, and Google piloting AI-assisted rounds, the AI-era technical round has arrived.

The candidates who fail this round fail in two ways: they freeze because “I’m not technical,” or they passively accept whatever the AI produces and can’t explain a single decision. The candidates who pass treat the AI like a talented junior engineer: clear direction, tight feedback loops, healthy skepticism.

Pro tip: This week, build one tiny thing with an AI tool: a prototype of a feature idea you’ve had, however scrappy. The goal isn’t the artifact. It’s crossing the line from “person who’s read about AI” to “person who’s shipped something with it,” because that difference is audible in an interview within about ninety seconds.

Share

Section 5: A note on figuring out which rounds you’ll face

One more thing before we zoom out, because it saves people from the exact blindside my student walked into: you’re allowed to ask.

When a recruiter reaches out or schedules your loop, ask directly: “Can you walk me through the interview stages and what each one covers? Will any round involve working with AI tools live?” Recruiters answer this question all day; it signals preparation, not weakness. Companies that have redesigned their process generally want you to know. Canva published a whole blog post about theirs, and Meta briefs candidates on the AI-enabled environment before the interview. Beyond asking, check the company’s official careers or engineering blog (that’s where these announcements live), read recent candidate reports for your specific role and level, and re-read the job description with fresh eyes. Phrases like “AI-native,” “prototype with AI tools,” or “reflexive AI usage” are not decoration. They’re telling you what the interview will sample.

Prepping blind is how strong candidates lose to average ones who simply knew the format. Five minutes of asking removes the single biggest source of interview-day surprise.

Section 6: The thread running through all seven

Look back across the rounds and you’ll notice they’re all sampling the same underlying skill from different angles: can you think out loud, with judgment, using your own evidence, with AI when it helps and without it when it matters?

That’s why the “there’s so much to study” panic (and I hear it on almost every call) is misdirected. You don’t have seven different exams to cram for. You have one skill expressed seven ways, and it’s built on two foundations.

Positioning. Knowing exactly what you can do, with the receipts organized: resume, LinkedIn, and a bank of deep stories tagged by competency, so the right evidence is at your fingertips in any round.

A working AI system. Not thirty bookmarked tools, but one setup with your context loaded in, used daily enough that reflexive AI usage becomes true of you before anyone tests for it.

Everything else (the frameworks, the question banks, the tips) hangs off those two or falls apart without them.

Part 7: Next steps

Everything in this newsletter is doable on your own. I want to be completely straight about that, because I have no interest in manufacturing helplessness to sell you something. So let me lay out the full do-it-yourself path (the real one, with real time estimates from watching dozens of people attempt it) so you can decide with clear eyes.

📌 The do-it-yourself (DIY) path

Step 1: Rebuild your positioning (15–20 hours). Pick your three target job descriptions. For every major requirement, find the evidence in your career: the specific thing you did, the outcome, the artifact. Then rewrite your resume so each bullet closes the chain from requirement to proof. This is not a wordsmithing exercise; it usually means archaeology through old docs, launch decks, and performance reviews to recover what you actually did, because most of us genuinely forget. Expect to throw away your first full draft. Nearly everyone does, because the first draft is still the old resume wearing new adjectives.

Step 2: Bring LinkedIn into alignment (4–6 hours). Headline as role + domain + proof. About section that a recruiter can skim in fifteen seconds and place you. Experience section that agrees with the resume; remember, they cross-check, and mismatches read as friction or worse. Then the part people skip: activity. A profile that shows you engaging with your target domain reads very differently from one that’s been dormant since your last job search.

Step 3: Build your story bank (10–12 hours). Inventory the 6–10 defining moments of your career. Write each one out longhand (context, what you did, what happened, what you’d do differently), then tag each by the competencies it can carry: execution, conflict, ambiguity, influence, judgment. This is the single highest-leverage prep asset you’ll create, and it’s also where working alone hurts most, because you can’t easily tell which of your stories actually land. What feels impressive from inside your own career is frequently not the thing an interviewer finds impressive, and vice versa.

Step 4: Study the interview stack (10–15 hours, ongoing). Map which of the seven rounds your target companies run: ask recruiters, read their engineering blogs, check recent candidate reports. Then drill the rounds you’ll actually face. For product sense and analytical: out-loud practice, recorded, reviewed. Painful and non-negotiable. For execution: build your metric trees. For behavioral: run your story bank against common probes and, more importantly, against follow-ups.

Step 5: Build your AI system (6–10 hours to set up, then daily use). One workspace (a Claude project or equivalent) with your resume, your story bank, and your target JDs loaded in, so every output starts from your evidence instead of a blank page. Use it to map your experience against each JD, generate practice questions specific to your gaps, and run mock rounds. Then use AI to build one real thing, however small, so the vibe-coding round finds you already fluent. The setup is the easy part; the habit is the point. “Reflexive AI usage” can’t be crammed the night before.

Step 6: Apply with a plan (4–6 hours to build, then weekly discipline). A target list of 20–30 companies tiered by fit, a weekly application cadence tied to your positioning work, and a tracking system so you can see which version of your story gets responses, because your applications are themselves an experiment, and most people never instrument it.

Add it up and you’re looking at 50–70 focused hours. Spread across evenings and weekends, alongside a job or an active search, that’s realistically 8 to 12 weeks: a full quarter, if you execute cleanly, in the right order, without stalling.

Where the solo path actually breaks down

And here’s the part I have to tell you, because I’ve watched this movie many times: it’s almost never the content that defeats people. It’s four quieter things.

1. Under stress, people do the steps in exactly the wrong order. They apply first (because silence feels like a volume problem), practice interviews second, and fix positioning last, if ever. Which means their strongest applications go out after they’ve burned through their best-fit companies with their weakest materials.

2. No feedback loop. Alone, you can’t hear yourself. You can’t tell if your positioning is sharp or just newly worded, whether your stories land or ramble, whether your product sense sounds like judgment or a framework in a trench coat. So people polish blind (the third full resume rewrite with no external signal is almost a rite of passage), and polishing blind mostly relocates the problem.

3. Somewhere around week five of doing this alone, the initial energy is gone, the applications sent with the old materials have returned their silence, and the new system isn’t finished yet. This is the exact point where most solo attempts quietly end. Not with a decision, but with a week off that becomes a month.

4. Isolation. Job searching is lonely in a way that compounds. Every rejection is processed alone, every small win has no witness, and there’s no one to tell you that the thing you’re stuck on is normal and survivable. I believe this, not the resume and not the frameworks, is why capable people stay stuck.

None of these are character flaws. They’re structural features of doing hard, ambiguous work alone with no deadline. Which is precisely why I stopped telling coaching clients “just follow the steps” years ago.

(starts Aug 1st, this weekend)

The Land Challenge is that entire path, same steps and same order, compressed from a solo quarter into four weeks, live with me, with the four failure modes designed out of it.

Here’s how the weeks map to what you just read:

Week 1: Position yourself. We rebuild your resume and LinkedIn around what you actually did: targeted for the job, specific to you. That’s Steps 1 and 2, with my eyes on your positioning instead of you guessing in a vacuum, plus we start rebuilding your career stories together.

Week 2: Learn the new AI PM interview stack. All seven rounds from this newsletter, round by round: what each one really tests, the psychology behind the questions, and how to answer. That’s Step 4 without the folklore: you prep for the test that exists.

Week 3: Prepare with AI. You build your own Claude project with your resume, stories, and target roles loaded in, so every prep session is specific to you, not a generic prompt. The exact prompts and practice loops I use with my own coaching clients, walked through step by step. That’s Step 5, done in a day instead of abandoned in a tab.

Week 4: Apply with a plan (except that we start with this week 1) A job search strategy that’s custom to you: which jobs are worth your time, how to use AI to move faster without sounding like everyone else, and the cadence to run after the cohort ends. Step 6, instrumented properly.

And the structural stuff the solo path can’t give you: every session is live, so there’s a real deadline each week; the discouraging middle never gets room to open up. Every session is recorded and the slides and materials are yours to keep, so missing a live call costs you nothing. And you do the whole thing inside a small cohort of people at the same stage doing the same work, which is the quiet engine of the entire thing. Watching someone else workshop their positioning teaches you things about your own that no article can.

The Challenge is $499, one-time. We start Saturday, August 1, and enrollment closes when the cohort fills or Saturday morning, whichever comes first.

Two honest filters before you click, because I’d rather you self-select out than join for the wrong reasons. This is not a magic wand. These are real strategies that work if you put in the work, and if you’re expecting interview calls to materialize overnight, save your money. And it’s not 1:1 interview prep; if you need individualized coaching, reply to this email and we’ll talk about that instead.

And my promise, in plain words: I can’t guarantee you a job. No one can, and you should be suspicious of anyone who says otherwise. But show up and do the work, and if you don’t walk away with clearer positioning and a better understanding of AI PM interviews, get in touch with me. I’ll refund you.

📌📌 [Join the Land Challenge → we start Saturday]

Either way, whether you join or take the solo path with the steps above, please stop prepping for the 2021 interview. The companies wrote the new rules down in public. Now you’ve read the memo.

Rooting for your search,

Nazuk

P.S. If you know one PM who’s been applying into silence for months, forward them this.

Sources: Shopify: Tobi Lütke’s AI memo · eMarketer on Shopify’s AI hiring gate · Canva Engineering: “Yes, you can use AI in our interviews. In fact, we insist.” · Meta interview process 2026 (AI-enabled coding round) · Google’s AI-assisted coding interview · Anthropic: Guidance on Candidates’ AI Usage · Axial Search: AI product jobs 2026 analysis

That is where many strong candidates go wrong. They answer with their biggest achievement, their most polished launch, or the story they practiced most. The interviewer may be looking for something else entirely: conflict resolution, customer judgment, operating rigor, learning from failure, or leadership at the level of the role.

Sales psychology would describe this as selling to the buyer’s need. In an interview, the requirement is simple: understand the hiring need, then present the strongest truthful evidence that addresses it.

Do not invent a different fruit. Do not force an orange into every answer. Choose the right evidence from the work you have actually done.

Behavioral interviewing is not about telling the interviewer everything that makes you impressive. It is about making your decision-making, ownership, collaboration, impact, and learning visible in a form the interviewer can evaluate.

Throughout this newsletter, I will show you how to identify what the interviewer is testing, choose the right stories, structure answers without sounding robotic, survive detailed follow-up questions, and calibrate your evidence for the company, role, and level in front of you.

Table of Content

00 Introduction: The Lemon Problem

01: What Behavioral Interviews Actually Evaluate

02: Why Strong Candidates Fail

03: Company, Role, and Level Calibration

04: Build Your Evidence Bank

05: STAR Upgraded: Headline, Judgment, and Learning

06: The Follow-Up Gauntlet, the Qs interviewer is supposed to ask to probe more

07: Interviewer Psychology and Bias

08: Tactical Moves and Rescue Scripts

09. Difficult Questions and Red-Flag Topics

10: Annotated Sample Answers by Seniority

11: A 60-Minute Behavioral Interview Walkthrough

12: The 48-Hour Speed Prep

13: Self-Scoring and Post-Interview Debrief

14: Toolkit, Resources

Section 01: What Behavioral Interviews Actually Evaluate

Behavioral questions use the past as evidence. The interviewer is not asking for a story because stories are entertaining. They are trying to predict how you may operate when a similar problem appears in the new role.

1.1 The Three Questions Behind Most Behavioral Questions

1. What was difficult or important about the situation?

2. What did this candidate personally notice, decide, and do?

3. Does this evidence predict success at the scope and level we are hiring for?

A good answer reduces the amount of inference required. It gives the interviewer enough context to understand the stakes, enough action detail to distinguish your contribution from the team’s, and enough reflection to show that the lesson changed how you operate.

1.2 Signal, Evidence, and Risk

Signal: The competency the interviewer is trying to assess, such as ownership, judgment, collaboration, customer focus, analytical rigor, or learning agility.

Evidence: The specific decisions, behaviors, outcomes, and lessons from your example that support the signal.

Risk: The concern the interviewer is trying to rule out, such as blame, weak ownership, shallow thinking, poor stakeholder judgment, or inability to operate at the target level.

EXAMPLE Question: Tell me about a conflict. Signal: productive collaboration. Evidence: you understood the other person’s incentives, changed the conversation, and reached a durable decision. Risk: you treat disagreement as a battle to win.

1.3 What the Interviewer Eventually Has to Write

In a structured process, the interviewer may need to submit written feedback against assigned competencies. Imagine the note they should be able to write after your answer:

Strong evidence of judgment. She identified the retention risk, rejected a faster but fragile option, aligned engineering and support around a two-week test, and improved 90-day retention by 18 percent.

If your answer makes a note like that easy to write, you have probably delivered usable evidence. If the note would be vague, such as smart person who worked on an interesting launch, the story may be impressive but the signal is not yet clear.

Section 02: Why Strong Candidates Fail

Failure Mode 1: You Prepared Scripts, Not Evidence

A memorized answer can collapse when the interviewer changes the wording, interrupts, or probes a detail you did not rehearse. It can also sound polished but lifeless because your attention is on recall, not conversation.

Prepare the bones of the story instead: the stakes, your decision, two or three actions, the outcome, and the lesson. Practice different ways into the same evidence.

Failure Mode 2: The Story Answers a Different Question

Candidates sometimes hear one familiar word and launch into a prepared story. A question about changing your mind is not automatically a data question. It may be testing humility and belief updating. A question about conflict may be testing relationship repair, not persuasion.

Failure Mode 3: Your Role Is Unclear

Team achievements matter, but the interviewer must still isolate your contribution. Too much, we can make a strong project sound like borrowed impact. Too much, I can make a collaborative project sound self-centered or implausible.

Failure Mode 4: Your Example Does Not Match the Level

A well-told low-scope story may still be weak evidence for a senior role. Conversely, an early-career candidate should not imitate executive language. The interviewer wants credible performance at the next level, not inflated scale.

Failure Mode 5: You Report Activity Instead of Judgment

I scheduled meetings, gathered requirements, worked with design, and launched the feature describes motion. It does not show why you chose the path, what you ruled out, or what changed because of your judgment.

Failure Mode 6: You Protect Yourself Too Much

Difficult questions are often risk checks. Defensiveness, blame, vague language, and fake perfection create more concern than a real mistake handled with maturity. Strong candidates can be accountable without turning one event into a confession or an identity statement.

Failure Mode 7: You Ignore Delivery

Excellent content can become hard to score when it arrives as a six-minute monologue. Aim for a clear first pass of roughly two to three minutes, then let the interviewer decide where to go deeper. Exact timing varies, but disciplined brevity creates room for the evidence they care about most.

Subscribe now

Section 03: Company, Role and Level Calibration

The same story can be strong for one role and insufficient for another. Calibration has three layers: the company’s evaluation system, the function’s core work, and the scope expected at the target level.

1: Company Context

2: Role Context

3: Level Context

Questions to Ask the Recruiter

1. How many rounds are in the process, and how long is each one?

2. Which competencies or values are assigned to the behavioral round?

3. Is the interview entirely behavioral, or is it combined with product, execution, or technical questions?

4. What level is the role calibrated for, and what scope would the new hire own?

5. Who will I meet, and what are their functions relative to this role?

6. Is there anything candidates commonly misunderstand about this process?

You are not asking for the answers. You are asking for the evaluation context so your preparation matches the actual job.

Section 04: Build your Evidence

Most traditional preparation starts with a list of questions. Better preparation starts with your career evidence. Questions change. Your experiences are the raw material that remains.

4.1 Operating from memory fails under pressure

Nobody teaches you to document a career. You spend years doing the work, then try to reconstruct launches, conflicts, decisions, numbers, and lessons a few days before an interview. Stress makes retrieval harder at exactly the moment you need it most.

I learned this personally after returning from maternity leave. Projects I had led for months felt fuzzy. I had just totally forgotten the metrics I used to remember, top of my head, all the time, the projects that I was leading. I had forgotten a lot of stuff.

Share

4.2 Where to Mine Stories

1. Performance reviews, promotion packets, resumes, and brag documents.

2. Roadmaps, strategy documents, launch notes, decision memos, and retrospectives.

3. Dashboards, experiment readouts, customer research, and support themes.

4. Emails or messages that capture a hard decision, feedback, escalation, or thank-you.

5. Moments when a plan failed, a relationship became difficult, or your belief changed.

6. Work that improved a system or another person’s performance, even if it did not produce a launch.

4.3 Techniques for good memory

4.4 Results Without Perfect Metrics

Not every meaningful story has a clean revenue number. Use the strongest honest evidence available, and use approximate ranges rather than invented precision:

22. Business or customer outcome: revenue, retention, adoption, conversion, satisfaction, or risk reduction.

23. Operational outcome: cycle time, quality, reliability, cost, throughput, or fewer escalations.

24. Decision outcome: a roadmap changed, a bet stopped, an investment approved, or a risk avoided.

25. Behavioral evidence: repeated adoption, stakeholder commitment, customer quotes, or a durable process change.

Section 5: STAR upgraded: headlines, Judgement and learning

5.1 Start With a Headline

The headline gives the listener a map. The details that follow have somewhere to land.

5.2 Compress Situation and Task

Include the product, goal, stakes, constraints, and your role. Leave out organizational history unless it changes how the decision should be understood.

We were losing 60 percent of trial users at step three. I owned activation, and we had six weeks before the next board update to determine whether the issue was motivation or comprehension.

5.3 Expand Actions Into Decisions

A list of tasks is not enough. Use the action section to answer questions such as:

1. What did you notice that others had not yet connected?

2. Which alternatives did you consider, and why did you reject them?

3. What tradeoff did you accept?

4. How did you earn commitment from people with different incentives?

5. What changed after new evidence appeared?

6. Which action was specifically yours?

5.4 Make the Result Credible

Results need a baseline and an attribution boundary. If the team delivered the outcome, say we. If your decision changed the path, explain that contribution. Connect the metric to customer or business meaning with a so-what sentence.

Activation improved by 11 points, and the change became our second-largest source of new retained users that quarter. I led the diagnosis and recommendation; design and engineering owned the implementation with me.

5.5. End With Changed Behavior

A lesson is strongest when it changes a later action. Avoid generic endings such as communication is important.

5.6 The I and We Rule

Simple Rule: Use I for decisions, reasoning, and actions you personally owned. Use we for shared execution, outcomes, and credit. Strong answers can do both without shrinking the team or disappearing inside it.

5.7 Full Answer

Share

Section 06: The Follow-Up Gauntlet, the Qs interviewer is supposed to ask to probe more

Strong interviewers often spend more time on follow-up questions than on the initial prompt. That is not automatically a sign that your answer failed. They might be testing ownership, factual consistency, judgment, learning, or level.

6.1 The Ten Follow-up Qs You Should Expect

1. What exactly was your role?

2. Why did you choose that option?

3. What alternatives did you consider?

4. Who disagreed, and why?

5. What did you say or do to change the situation?

6. How did you measure success?

7. How do you know your action caused the result?

8. What went wrong along the way?

9. What would you do differently now?

10. Where did you apply that lesson later?

6.2 Why Candidates Unravel

This particularly happens when you might have memorized the surface story but not really understood the structure behind it. I wrote an exact post here

If you use AI to write your stories, you must instruct the AI not to make up facts or sound above your experience level. It has to sound at your experience level.

The language sometimes makes us think, “Wow, it’s pretty nicely written,” but when you go into the interview, the interviewer is not looking for a perfect answer or perfect memory. They are looking for credibility and depth, and they are looking at your level. If you have three to five years of experience, your answers should match that

6.3 The Follow-Up Drill

1. Deliver the story in two to three minutes.

2. Have a partner choose five random probes from the list above.

3. Answer each probe directly in 20 to 45 seconds before adding detail.

4. Check whether the timeline, ownership, and metrics remain consistent.

5. Repeat with a different opening question that maps to the same story.

6.4 Useful Scripts Under Pressure (I use this technique often)

This allows you to ask a clarifying question to make sure you’re answering the right one. It also gives you a little bit of time to not have a cold start, meaning right when the question is asked, you don’t directly jump into the question. When you further ask a follow-up question to the interviewer, it also helps you calm your own nerves and buys you a little bit of time.

1. If you need clarification: When you say conflict, are you most interested in how I reached the decision or how I repaired the relationship afterward?

2. If you do not remember the exact number: I do not remember the exact figure, but it was in the 15 to 20 percent range. The dashboard showed a clear improvement over the prior cohort.

3. If causality is imperfect, I cannot attribute the full change to my decision because two other releases shipped that month. The cleanest evidence was the controlled segment, where we saw an eight-point lift.

4. If information is confidential, I cannot share the absolute revenue figure, but I can describe the decision, the relative change, and how we validated it. Does that work?

5. If the premise is slightly off: The closest example is not a direct report conflict. It was a cross-functional disagreement with similar incentives and stakes. Would that be useful?

6.5 Watch for the Ownership Funnel

6.6 Depth Without a Data Dump

Keep a small set of proof details available: baseline, sample size, time frame, stakeholder incentives, rejected options, and what happened next. Offer the detail that answers the probe. Do not unload every fact because you are relieved to remember it.

Section 07: Interviewer Psychology and Bias

Interviewers try to be objective, but they are still human. Structured questions and defined scoring criteria can improve consistency, yet first impressions and cognitive shortcuts can still affect how evidence is noticed and interpreted.

Use this knowledge ethically. The goal is not to manipulate the interviewer. The goal is to reduce ambiguity and make job-relevant evidence easier to see.

1. Primacy Bias: Your Opening Sets the Tone

First impressions can anchor the conversation. A clear, relevant opening makes later evidence easier to interpret. A hesitant or unfocused start can create uncertainty that later answers must overcome.

2. The Halo Effect: One Strong Moment Can Color the Rest

A strong answer in an area the interviewer values can positively affect how other evidence is interpreted. The reverse is sometimes called the Horn Effect.

3. Confirmation Bias: Early Impressions Can Reinforce Themselves

After forming an initial view, people may listen more closely for evidence that confirms it. Specific detail and direct answers reduce room for unsupported inference.

4. Affinity Bias: Familiarity Can Be Mistaken for Fit

Interviewers can favor candidates who share their style or background. Do not imitate a persona. Focus on clarity, professional warmth, and evidence relevant to the job.

5. Recency Bias: Your Closing Evidence Stays Fresh

The final answer may be among the freshest evidence when feedback is written. Use it to fill a competency gap or reinforce an important signal.

7.1 The Kahneman Connection (I discovered this so late in my career)

Daniel Kahneman’s System 1 and System 2 framework is useful for understanding how fast impressions and deliberate evaluation can interact.

Sequence and framing matter alongside content. A headline helps the fast system orient, while decisions and proof give the analytical system something to score.

7.2 Do Not Confuse Fluency With Substance

Have you ever heard a person talk a lot of words, but when they’re done talking, you say, “You asked them, ‘What did you mean, really? What happened?’

You’re left even more confused: what were they trying to say, or what was the point? You’re just more lost because it sounds like a lot of nice words, but there’s no substance underneath.”

7.3 Make Bias Less Powerful

1. Answer the exact question before adding context.

2. Use concrete verbs and observable decisions.

3. State your role and attribution boundary.

4. Name the tradeoff and rejected alternative.

5. Close with impact and changed behavior.

6. Ask clarifying questions when the evaluation target is ambiguous.

Subscribe now

Section 08: Tactical Moves and Rescue Scripts

Some of these techniques I learnt over 15 years of my career

1: Lead With Your Strongest Relevant Story

Do not save your best relevant evidence for later. Relevance comes before scale. Choose a story that directly answers the question, carries meaningful stakes, and shows judgment.

2: Understand the Interviewer’s Professional Context

Review the interviewer’s public professional background, the team, and the product. Use that context to choose understandable vocabulary and relevant evidence. Do not use it to manufacture a persona.

3: Use the So-What Close

Do not stop at we shipped or retention improved. Add one sentence that explains the business, customer, or organizational meaning.

That feature became the second-largest driver of new-user activation that quarter, which changed how we allocated the next roadmap cycle.

4: Name the Tradeoff

Option X would have shipped two weeks earlier, but I chose Y because the faster path created a retention risk we could not reverse cheaply.

5: Manage the Signal Portfolio

Four questions are not four isolated answers. Together, they create a portfolio. Track which competencies you have already demonstrated and use later questions to add range.

6: Pause Before You Answer

A short pause gives you time to select the right evidence and signals that you are choosing deliberately. You can say, Let me take a moment to choose the best example.

7: Use a 90-Second Introduction

• Throughline: Name the problem, skill, or perspective that connects your career.

• Proof point: Use one specific result that makes the throughline credible.

• Forward hook: Explain why this role and why now.

Example: I have spent my career turning complex customer friction into simple product decisions. At my current company, I led an onboarding redesign that improved activation by 14 points by removing steps rather than adding features. I am now looking to apply that experience at a larger scale, which is why this team’s adoption problem caught my attention.

Section 09: Difficult Questions and Red Flag topics

9.1 This is critical to follow

1. State the difficult fact directly without dramatic buildup.

2. Own your contribution without absorbing blame that is not yours.

3. Explain the corrective action and outcome.

4. Show the behavior, system, or judgment that changed afterward.

9.2 How to share a failure Story (classic tell me about a time you failed)

Choose a real miss with meaningful stakes. Avoid catastrophic events that remain unresolved, tiny mistakes with no consequence, or disguised strengths such as I cared too much.

9.3 “Tell me about a time you received negative feedback”

Do not spend the answer proving the feedback was unfair. Explain what you heard, how you tested it, what you changed, and the evidence that the change stuck.

9.4 Conflict

A strong conflict answer explains the other person’s legitimate incentives. If the story requires the interviewer to believe everyone else was irrational, the risk may be you.

9.5 A decision that was wrong

Separate a reasonable process with a bad outcome from a poor process you need to own. In either case, explain what new evidence changed your belief and how quickly you responded.

9.6 Weakness

Choose a real, bounded pattern that is not fatal to the role. Show the mechanism you use to manage it and recent evidence of improvement. Do not claim the weakness has vanished permanently.

9.7 Layoff

Strong Answer:

My team was part of a broader restructuring after the company changed strategic priorities. My product area was deprioritized. It was not performance-related, and it gave me a reason to think carefully about what I want next. I am looking for a role where the product is a core company bet, which is what drew me to this team.

9.8 Career Gap or Caregiving Break

Strong Answer:

I took two years off to be the primary caregiver for my children. It was a deliberate decision. During that time, I stayed connected through an analytics course and pro bono product advisory work. I am now ready to return full-time, and I am targeting roles where my prior product experience and renewed analytical focus are directly relevant.

9.9 Fired or Managed Out

Use accurate language, keep context proportional, and do not attack the employer. Name the performance or fit issue, what you learned, and what you changed. If legal terms or agreements restrict what you can say, stay within them.

What Creates Concern

1. Blaming a manager, colleague, customer, or company for the entire outcome.

2. Using vague language that avoids the difficult fact.

3. Claiming a lesson without evidence that behavior changed.

4. Turning every weakness into a flattering trait.

5. Sharing confidential details or over-correcting into a long confession unrelated to the job.

Section 10: Annotated Sample Answers by Seniority

HOW TO USE THESE SAMPLES

Do not copy the wording. Copy the evidence pattern: direct answer, necessary context, visible decision, honest ownership, meaningful outcome, and changed behavior. Your natural voice is an advantage when the structure underneath it is sound.

10.1 Tell Me About a Time You Made a Decision Without Enough Data

Hiring manager reads: Independent judgment, a reversible way to reduce uncertainty, clear ownership, and measurable product impact.

Hiring manager reads: Wider scope, competing executive incentives, risk allocation, multi-team alignment, and a durable operating model.

10.2 Tell Me About a Time You Influenced Without Authority

Hiring manager reads: Problem discovery, initiative, stakeholder empathy, adoption, and product-relevant insight without a PM title.

Hiring manager reads: Influence through problem framing, incentive alignment, sequencing, and organizational leverage rather than escalation alone.

10.3 Tell Me About a Failure

Hiring manager reads: Credible early-career stakes, direct ownership, fast correction, and a concrete learning loop.

Hiring manager reads: A meaningful miss, no blame, proportionate recovery, and a change in strategic decision quality.

10.4 I See a Gap on Your Resume

Hiring manager reads:Direct, non-defensive, current readiness, and a clear return to job-relevant evidence.

10.5 Why Are You Leaving Your Current Role?

Hiring manager reads: Honest, free of bitterness, and connected to a credible reason for this role.

Section 11: A 60-minute Behavioral Interview Breakdown

Important: This is an illustrative 60-minute format, not a universal schedule. Many behavioral rounds are 45 minutes, combine functional questions, or spend most of the time probing only two stories.

11.1 60-minute Behavioral Interview breakdown

11.2 Don’t Over-Interpret

A neutral interviewer is not necessarily disengaged. Detailed probing is not necessarily negative. Running out of time is not automatically a rejection. Interview styles differ. Stay focused on the question, the evidence, and the conversation in front of you.

Section 12: The 48-hour speed prep

The 48-Hour Sprint

12.1 First couple of hours

• Read the job description and identify the five most likely competencies.

• Confirm the interview format and level with existing recruiter information.

• Choose six to eight stories that provide the best coverage.

• Flag any area with no credible evidence.

12.2 Next Three Hours

• Create one-page story cards.

• Verify the timeline, metrics, and attribution for each story.

• Write a one-sentence headline and one-sentence lesson for each.

• Prepare the 90-second introduction and four questions for the interviewer.

12.3 The Next Day

• Run one full mock interview.

• Drill the ten follow-up probes on the three most likely stories.

• Repair the weakest answer, then stop rewriting everything.

• Check the interview link, location, time zone, device, audio, and backup plan.

12.4 The Evening Before

• Review headlines and proof details, not full scripts.

• Lay out materials and remove avoidable logistics risk.

• Choose a reasonable stopping time. Sleep is part of retrieval performance.

12.5 Thirty Minutes Before

• Review the role’s top three signals.

• Recall the opening line of four core stories.

• Say your introduction aloud once.

• Close notes, breathe, and enter the conversation rather than another study session.

  IF YOU HAVE ONLY ONE HOUR:

  Choose three stories: one ownership win, one conflict, and one failure or feedback example. For each, write the headline, decision, result, and learning. Then practice two follow-ups per story.

12.6 Virtual Interview Setup

• Place the camera near eye level and test audio in the actual room.

• Keep story titles and interviewer questions visible, not scripted paragraphs.

• Silence notifications and close bandwidth-heavy applications.

• Keep water and a backup connection or dial-in available.

• If technology fails, name the issue calmly and move to the backup plan.

Section 13: Self-Scoring and Post-Interview Debrief

You can Use AI for sections 13.1 and 13.2.

You can use this sample prompt

13.1: Self-Scoring or use AI

13.2: How to Review a Recording

• First, listen for content only. Could an interviewer write specific evidence?

• Listen again for structure. How long before the answer became clear?

• Review ownership. Count where I and we were accurate or misleading.

• Review judgment. Did you name an alternative, tradeoff, or reversal criterion?

• Review delivery. Note pace, filler, long detours, and whether the ending lands.

• Choose one change for the next repetition. Do not rewrite the entire answer at once.

13.3 Quick Post-Interview Debrief

Complete this while the conversation is fresh. It turns every interview into better evidence for the next one.

Share

13.4 How to Interpret a Rejection

Hiring outcomes are noisy. Team needs, level calibration, competing candidates, headcount, and interviewer variance can all matter. A rejection is not a clean measurement of your worth.

Still, recurring patterns deserve attention. If three independent interviewers ask what you personally did, treat ownership as a practice priority. If feedback repeatedly mentions seniority, examine the scope and leverage of your evidence rather than simply polishing delivery.

13.5 When Feedback Is Vague

Translate vague feedback into observable practice targets. Executive presence might mean a missing headline, weak prioritization, or overlong context. Strategic depth might mean the answer lacked alternatives, market implications, or long-term tradeoffs. Ask a coach or peer to score the recording rather than guessing from the label alone.

Section 14: Toolkit, Resources

14.1 Books

📖 Thinking, Fast and Slow: Daniel Kahneman

The foundation for everything in Part 02 of this guide. System 1 vs. System 2 is the single most useful mental model for understanding how interviewers make decisions.

14.2 Videos Worth Your Time

▶ Simon Sinek: How Great Leaders Inspire Action

TED Talk · 60M+ views

The “Start With Why” framework. Learn to lead with purpose in your stories.

▶ Nancy Duarte: The Secret Structure of Great Talks

TED Talk

She found that the best talks contrast “what is” with “what could be.” Use this structure in your STAR stories.

▶ Julian Treasure: How to Speak So People Want to Listen

TED Talk · 50M+ views

Tone, pace, register: the vocal mechanics that make the difference between a flat answer and one that commands attention.

14.3 Your Ongoing Career System

• Update the evidence bank once a month or after every major decision.

• Save metrics, feedback, and artifacts while details are still accurate.

• Review coverage before performance cycles and interviews.

• Run one follow-up drill whenever you add a new story.

• Use post-interview notes to improve the next repetition.

REMEMBER, 

You cannot sell oranges to someone who needs lemons. Listen for the need, choose the most relevant truthful evidence, and make it easy for the interviewer to see you doing the work at the level the role requires.

If you made it this far and liked reading this newsletter, you will love this 👇👇👇

See you next week,

-Nazuk

And if you sound like everybody else, you disappear.

That is the biggest mistake I see people making right now. They open ChatGPT and say:

• “Rewrite my resume.”

• “Make this bullet point sound better.”

• “Make me sound more senior.”

• “Tailor this resume for this job.”

And then they copy-paste whatever comes out.

The result is usually something like this:

• “Led cross-functional initiatives to drive 15% engagement growth and improve stakeholder alignment across product, design, and engineering.”

• “Spearheaded data-driven strategies that increased operational efficiency by 22% while optimizing user-centric workflows.”

• “Owned end-to-end execution of strategic roadmap initiatives, resulting in measurable business impact and improved customer outcomes.”

Now, does this sound fancy?

Yes.

Does it sound professional?

Also yes.

But does it sound like you?

Probably not.

And more importantly: can you defend every word in an interview?

maybe,

Because sometimes ChatGPT writes your work in a way that makes you think, “Wow, I did that?” And maybe you did. Maybe kind of. Maybe not exactly like that.

That is the trap!!!!!

HI, Today’s newsletter will focus on just one thing: job search

• Job Application Strategies

• Anti-AI Strategies

• How many jobs should you be applying to land something in the next 90 days?

• The strategies that worked for me twice and for my husband while we were both applying to jobs at the same time. He is an engineer for context.

• Prompts that worked for me

• How to use Claude, Codex, and Perplexity Computer for job search.

Section 1:

See, I am not a native English speaker. My first language is Hindi. Even today, after years of building my communication skills, my brain still thinks in Hindi first.

So when I try to explain my work in English, sometimes I know exactly what I did, but I struggle to make it sound sharp, clear, and professional.

That is where ChatGPT feels magical.

You give it your rough sentence, and suddenly it gives you a polished bullet that sounds like it belongs on a McKinsey resume.

But polished does not always mean truthful. And polished does not always mean effective.

Your resume should not sound like AI wrote it. Your resume should sound like proof.

Proof that:

• You have done the work.

• You understand the problem.

• You can explain your impact.

• You can connect your experience to the role.

• You can talk about your work clearly in an interview.

So no, this article is not about asking ChatGPT to “make your resume better.”

That is not the strategy.

Also, I want to be very clear. Nobody can guarantee you a job. What I can do is share everything that worked for me.

Last fall, my husband and I were both applying for jobs at the same time. We both used AI very intentionally. We tested prompts, changed strategies, tracked what worked, and stopped doing what did not work.

Both of us landed jobs in less than 90 days.

Again, that does not mean this will happen for everyone. Your market, your background, your role, your timing, and your network all matter.

But I can tell you this: Using ChatGPT the right way changed how we approached the job search. Not as a Blind Resume writer. That is the difference.

Section 2: The Job Search Strategy

Before we even talk about resume prompts, let’s talk about strategy.

Because if your job search strategy is weak, even the best resume will not save you.

Right now, if you are actively looking for a job, you need to be smart about timing. We are in July, which means you have a real window before November and December slow things down.

You have July, August, September, and October to move with focus. This is not the time to casually apply once in a while and hope something happens.

Your strategy should be:

• Apply to jobs within 24 hours of the posting.

• Target at least 5 strong applications per day.

• Use LinkedIn to find jobs, not necessarily to apply.

• Apply directly on the company website.

• Create a profile on every company career portal.

• Prioritize referrals where you already have a real connection.

• Be specific about the roles, domains, and companies you are targeting.

And yes, if you apply to 100 companies, does that mean you may end up creating 100 different company profiles?

Unfortunately, yes.

That is part of the process.

If you are applying to Microsoft, create a profile on Microsoft’s career site. Do not only click “Easy Apply” and move on.

LinkedIn is useful for discovery. It is not always the best place to submit your application.

Use LinkedIn to:

• Search for new job postings.

• See who works at the company.

• Check if someone in your network can refer you.

• Understand the hiring team or business area.

• Track roles that were posted recently.

Then go apply on the company website.

Subscribe now

Section 3: Where to Search for Jobs

Do not depend on one platform.

You should have a small system that you check every day.

Start with:

• LinkedIn for discovery.

• Company career pages for direct applications.

• Greenhouse job boards.

• Ashby Jobs.

• Built In SF if you are looking for startup jobs.

• Specific company career portals for companies you are serious about.

The goal is not to apply everywhere.

The goal is to become very fast and very specific.

You want to find the right job quickly, understand whether it fits, tailor your resume, and apply before the posting is flooded.

Resource:

• LinkedIn Jobs
  Best overall for PM roles, recruiter outreach, referrals and identifying hiring managers. Use LinkedIn to discover the opening, then apply through the company’s career page whenever possible.

• Indeed
  One of the largest job databases. Useful for PM, technical PM, program management and roles at companies that may not be highly active on LinkedIn.

• Built In
  One of the best dedicated portals for product and technology jobs. Particularly useful for startups, fintech, SaaS and mid-sized tech companies.

• Wellfound
  Formerly AngelList Talent. Best for early-stage and growth-stage startup roles. Listings frequently include compensation, equity, company stage and founder information.

• Welcome to the Jungle
  Formerly Otta in the US. Strong for curated technology and startup opportunities, with recommendations based on your preferred role, industry and company environment.

• Y Combinator – Work at a Startup
  Excellent for product roles at YC-backed startups. Especially valuable for founding PM, first product hire, AI PM and highly entrepreneurial product roles.

• Dice
  Focused specifically on technology careers. Useful for technical product manager, platform PM, data, AI, cloud, cybersecurity and contract opportunities.

• Glassdoor Jobs
  Best used for researching salaries, interview experiences and company reviews alongside the job listing. Verify that the position is still active on the company website before applying.

• FlexJobs
  Strong for remote, hybrid and flexible technology roles. FlexJobs screens its listings, although full access generally requires a paid membership.

  Activate 5% Cash Back

• Ladders
  Focused primarily on positions paying $100,000 or more. Useful for senior PM, principal PM, group PM and product leadership searches.

• ZipRecruiter
  A broad job aggregator with helpful alerts and employer-matching features. It may contain duplicates, so confirm every role on the employer’s official career page.

• USAJOBS – Tech
  The official portal for US federal government technology positions. Search for digital services, IT product management, program management, data and technology modernization roles.

• MyGreenhouse
  Greenhouse’s job-seeker platform for discovering and tracking openings from companies that use Greenhouse. It is particularly useful because many technology companies publish roles through Greenhouse before they spread across aggregators.

• Ashbyhq Jobs: This is an ATS
  Ashby hosts career pages for many startups and fast-growing technology companies. Search Google using queries such as:

site:jobs.ashbyhq.com "Product Manager" "United States"
site:jobs.ashbyhq.com "AI Product Manager"
site:jobs.ashbyhq.com "Senior Product Manager" "Remote"

Section 4: Big Tech, Mid-Sized Companies, and Startups Need Different Strategies

Do not use the same strategy for every company.

If you are targeting big tech or FAANG-type companies, referrals matter a lot.

For big tech, your strategy should be:

• Build a focused company list.

• Find people in your existing network first.

• Ask for referrals before applying.

• Apply through the company career portal.

• Make sure your resume clearly matches the role.

• Prepare for a longer interview process.

If you are targeting mid-sized companies, there may be less competition compared to big tech. Referrals still help, but direct applications can work too.

For mid-sized companies, your strategy should be:

• Apply directly on the company website.

• Create your candidate profile.

• Look for hiring managers or team members on LinkedIn.

• Send a short, specific message after applying.

• Move quickly when a role is posted.

If you are targeting startups, you need to show clarity and relevance very quickly.

For startups, your strategy should be:

• Be specific about why that company.

• Show domain knowledge.

• Show that you can work with ambiguity.

• Apply quickly.

• Reach out directly if there is a clear connection.

• Use platforms like Built In SF, Ashby, and Wellfound-style startup boards.

Share

Section 5: Referrals: Start With People You Already Know

When people hear “get a referral,” they immediately think they need to message strangers on LinkedIn.

That is not where I would start, especially if you are in a time crunch.

Start with your existing network.

Look at:

• Your former coworkers.

• Your classmates.

• Your managers.

• Your friends.

• Your alumni network.

• People you have worked with on projects.

• People you have had real conversations with before.

Then ask:

• Where do they work now?

• Is their company hiring?

• Are there roles that match your background?

• Can they refer you or point you to the right team?

My husband and I had different experiences here.

I applied mostly without referrals. My husband applied with referrals. He had people internally submit his resume. And honestly, he got more interviews lined up than I did.

He is an engineer and I am in product, so the roles were different. But the signal was clear: referrals helped.

Section 6: Do Not Go Broad on Role Fit

This is important.

When you are seriously job searching, this is not the time to apply to every role that sounds vaguely related to your background.

Do not look at a posting and say:

“Oh, they mentioned analytics. I have done analytics. Maybe I can do this job.”

That is too broad.

Be specific about:

• The work you have actually done.

• The domain you understand.

• The problems you can solve.

• The roles where your experience clearly maps.

• The companies where your background gives you an advantage.

For example:

• If a job says 2 to 4 years of experience and you are a fresh graduate with strong projects, apply.

• If a job says 8 years of experience and you have 5 strong years, apply.

• If a job asks for a domain you understand deeply, apply even if the title is slightly different.

But if the role is completely outside your experience and you cannot build a clear story for why you fit, skip it.

Section 7: How to Cross-Apply Your Experience

There is one place where you can be flexible: domain transfer.

You may not have worked at the exact same type of company, but you may have worked on the same kind of problem.

For example, let’s say you have experience in payments.

You could apply that experience in multiple ways:

• A payments company like Stripe, PayPal, Adyen, or Block.

• A marketplace company with a payments team.

• A fintech company building checkout or billing flows.

• A SaaS company with subscription payments.

• An e-commerce company optimizing checkout conversion.

• A creator platform handling payouts.

• A banking product with transaction workflows.

The company may be different, but the problem space overlaps.

You can build a story around:

• Reducing payment failures.

• Improving checkout conversion.

• Managing fraud and risk.

• Building billing or subscription workflows.

• Improving payout reliability.

• Understanding compliance and user trust.

• Working with engineering, design, data, and operations on payment flows.

That is cross-application.

You are not randomly applying. You are saying:

“I understand this problem space, and I can apply that knowledge in a new context.”

That is very different from applying broadly with no story.

Subscribe now

Section 8: Resume Strategies

Read rest of the sections by becoming part of my newsletter community.

Step 1: Create Your Skill Map + prompt

Step 2: Stop Applying With One Generic Resume + prompt

Step 3: Make Your Resume Sound Like You + prompt

Section 9: For Fresh Graduates: Your Strategy Has to Be Different

Section 10: What Did Not Work for me

Section 11: Final word/ the strategy

Section 12: AI Tools ++ Prompts for daily scheduled job search

Download

Read more

My core premise here is that

Most people use AI like a generic prompt box. Building your own Career “operating system” turns AI into a personalized career intelligence system by giving it your career evidence, target role, constraints, decision rules, and quality standards has personally helped me not only prepare for interviews, but also update my LinkedIn, position myself for the right opportunities, understand my key strengths, and so much more.

It even pulled a career arc I didn't know I had…

For example i have worked in B2B enterprize companies and it was very natural to have myself position as enterperize producr leader and as someone who understands nuances in large matrix organization but with this system I was able to understand that…

actually I can position myself as a 0→1 product leader too because I have led several initiaves from ground up and scaled it.

So today I am sharing what you will need to build it in ChatGPT or Claude, whichever is your AI of choice. I like Claude more. But I am actually going to try to build it in ChatGPT this week too and will report back.

What this guide helps you build

Use this as a setup guide, teaching asset, and copy-paste prompt library.

Table of Contents

1. Why Career OS Exists

2. What Career OS Actually Is

3. Why Projects Work Better Than One-Off Chats

4. The Career OS Architecture

5. What to Upload Into Your Career OS

6. How to Write Project Instructions

7. The Core Workflows

8. The Best Way to Work Inside Projects

9. Common Mistakes to Avoid

10. 7-Day Career OS Setup Plan

11. Copy-Paste Prompts

12. Sources Consulted

1. Why this system is useful

The biggest mistake product managers make with AI career tools is expecting a generic prompt to produce a personalized answer. They ask ChatGPT or Claude to write a resume, create interview answers, or improve a LinkedIn profile without giving the AI enough context to understand their career, target role, proof points, constraints, or positioning.

That is why the output often sounds polished but shallow. It may look professional, but it does not feel specific. It misses the nuance of a person’s background. It may overuse generic phrases like strategic product leader, cross-functional collaborator, or data-driven problem solver. Those phrases do not create differentiation because thousands of candidates use them.

Building your own Career OS solves this problem by turning AI from a one-off writing assistant into a structured career system. Instead of asking AI to guess, you give it a reliable source of truth. Instead of asking for a resume from scratch, you ask it to map your actual evidence to a specific job description. Instead of practicing random interview questions, you ask it to build preparation around your target role, your experience, and the stories you can actually defend.

The principle

Generic AI inputs create generic career assets. Specific career context creates specific, targeted, and more useful career assets.

For product managers and aspiring PMs, this matters because PM hiring is highly context-driven. A PM resume for a fintech platform role should not read the same as a PM resume for a consumer AI role. A story for a senior PM execution interview should not sound like a story for an associate PM behavioral round. A LinkedIn profile for AI product roles should emphasize a different signal than a profile for traditional SaaS product roles.

Career OS helps you build that context once, reuse it repeatedly, and improve it over time.

2. What Career OS Actually Is

Career OS is a personalized career operating system built inside a project workspace in ChatGPT, Claude, or a similar AI tool. It contains your career evidence, target roles, instructions, templates, interview stories, resume versions, and feedback loops.

It is not just a folder. It is not just a prompt. It is not just uploading a resume and hoping for better answers.

A strong Career operating system has four parts:

When these layers work together, AI can answer questions with more precision. It can identify which stories matter for a role, which bullets should be rewritten, which skills are missing, which achievements deserve more visibility, and where your positioning is weak.

Subscribe now

3. Why Projects Work Better Than One-Off Chats

Projects are useful because career work is not a single prompt. It is a long-running system. You need to return to the same context again and again: your resume, your stories, your metrics, your target companies, and your positioning.

As of June 2026, official guidance from OpenAI describes ChatGPT Projects as a way to group chats and files, add custom instructions, and keep work organized around a specific context. Official Anthropic guidance describes Claude Projects as self-contained workspaces with chat histories, project knowledge, and project instructions.

The platform details can evolve, but the operating principle is stable: put durable career context into the project, then run separate focused chats for specific workflows.

4. The Career OS Architecture

The easiest way to teach Career OS is to describe it as a command center with five modules.

This structure matters because every career asset is downstream of the same question: What are we trying to prove, and what evidence do we have? A targeted resume is not a creative writing exercise. It is evidence selection. Interview prep is not memorizing perfect answers. It is story retrieval. LinkedIn positioning is not personal branding fluff. It is choosing the right signal for the audience you want to attract.

5. What to Upload Into Your Career OS

Start with the assets below. You do not need all of them on day one, but the more complete your truth layer becomes, the more specific the AI output becomes.

The most important file is the brag document. Resumes show the polished version of your career. Brag docs show the raw material: the messy problems, decisions, numbers, tradeoffs, and moments that actually prove seniority.

Share

6. How to Write Project Instructions

Project instructions are the behavior layer of Career OS. They tell AI how to think, what to prioritize, what to avoid, and what quality standard to follow.

Good instructions should answer five questions:

1. Role: What role should AI play? Career strategist, resume editor, interview coach, recruiter, hiring manager, or product mentor?

2. Audience: Who is the output for? Recruiter, hiring manager, interviewer, LinkedIn audience, or promotion committee?

3. Evidence rules: Can AI infer, or should it only use uploaded proof? What should it do when evidence is missing?

4. Quality bar: Should output be concise, senior, metrics-driven, direct, ATS-friendly, or interview-ready?

5. Workflow: Should AI diagnose first, ask for the JD, map evidence, then write?

The best instruction is not vague motivation. It is an operating contract.

Here’s exactly how it works:

Below you will find my best copy-paste project instructions and prompts to build a LinkedIn profile and interview prep, and other sections continue.

7. The Core Workflows

8. The Best Way to Work Inside Projects in Chatgpt or Claude

9. Common Mistakes to Avoid

10. 7-Day Career OS Setup Plan

11. Copy-Paste Prompts

12. Sources Consulted

In Claude, you go to Projects → Project Instructions. In ChatGPT, it’s Custom Instructions inside a Project. You write one cohesive paragraph or block — not five separate entries — that weaves all five elements together naturally.

It looks like this in practice:

Project Instruction prompt 1:

Use the prompt above or the one below. Not together.

Try both, see which one brings better results for you.

Project Instruction prompt 2:

7. The Core Workflows

Once the project is set up, do not use it randomly. Use repeatable workflows. This is what turns Career OS from a repository into an operating system.

Workflow 1: Targeted Resume Creation

Use this workflow every time you apply to a role. The goal is not to create a pretty resume. The goal is to make the resume speak directly to the role’s hiring signals while staying truthful to your actual experience.

1. Paste the job description.

2. Ask AI to extract the role’s core hiring signals.

3. Ask AI to map your career evidence to those signals.

4. Ask AI to identify gaps, weak proof, or missing keywords.

5. Generate a targeted resume draft.

6. Ask AI to explain what changed and why.

7. Manually review for truth, accuracy, and voice.

Workflow 2: Interview Prep

MY BIGGEST HACK: Interview prep becomes stronger when AI can access your story bank. Instead of asking for random questions, ask AI to predict likely questions from the job description and match them to your strongest stories.

• Create a question bank based on the role, company, and seniority level.

• Map each question to one or two relevant stories from your story bank.

• Rewrite stories in a concise structure such as FAST or STAR.

• Run mock interviews and ask for feedback on clarity, seniority, specificity, and business impact.

• Create a final interview prep sheet with role signals, must-use stories, and questions to ask the interviewer.

Workflow 3: LinkedIn Positioning

Career OS can help you turn your background into a stronger public narrative. For PMs, this is especially useful when trying to signal a transition into AI product, platform product, fintech, consumer product, or leadership roles.

• Rewrite your headline around the role you want, not only the job you had.

• Use your About section to connect your experience, proof, and target direction.

• Turn product launches or lessons learned into LinkedIn posts.

• Use your portfolio or proof-of-work to make your profile more credible.

Workflow 4: Growth and Promotion

Career OS is not only for job seekers. PMs can use it to prepare for performance reviews, promotion conversations, manager check-ins, and career planning.

• Convert weekly wins into a monthly impact summary.

• Turn project work into promotion-ready narratives.

• Identify gaps between your current level and next-level expectations.

• Prepare concise updates for managers and leadership stakeholders.

8. The Best Way to Work Inside Projects in ChatGPT or Claude

The best Career OS setup is simple enough to maintain but structured enough to produce consistent output.

9. Common Mistakes to Avoid

10. 7-Day Career OS Setup Plan

You can set up a strong first version of Career OS in one week. The goal is not perfection. The goal is to create a usable system that improves with every application, interview, and career reflection.

11. Copy-Paste Prompts

Below this line, you will find my best copy-and-paste prompts for this setup!

Use the prompts below inside your Career OS project. The prompts assume you have uploaded your resume, LinkedIn, brag document, story bank, and other career assets into the project. If you have not uploaded them yet, the AI should ask you to provide them before generating final output.

Prompt 1: General Career Asset Prompt

Prompt 2: Targeted Resume from a Job Description

Prompt 3: Resume Fit Score + High-Impact Rewrite

12. Sources Consulted

The platform-specific notes in this guide were checked against official documentation. Product features can change, so always verify current settings inside your own ChatGPT or Claude account.

OpenAI Help Center: Projects in ChatGPT: https://help.openai.com/en/articles/10169521-projects-in-chatgpt

Anthropic Support: What are projects?: https://support.anthropic.com/en/articles/9517075-what-are-projects

Anthropic Support: How can I create and manage projects?: https://support.anthropic.com/en/articles/9519177-how-can-i-create-and-manage-projects

Anthropic Support: Understanding Claude’s personalization features: https://support.anthropic.com/en/articles/10185728-understanding-claude-s-personalization-features

Anthropic Support: Retrieval augmented generation for projects: https://support.anthropic.com/en/articles/11473015-retrieval-augmented-generation-rag-for-projects

Final reminder: Career OS does not replace your judgment. It gives you a better system for turning your real career evidence into targeted, role-specific career assets.

If you’ve been job searching recently, you’ve probably experienced this.

You spend an hour tailoring your resume. You answer all the application questions. You hit submit. Then at 2:07 a.m. on a Saturday, you receive an email:

“Thank you for your interest. We have decided to move forward with other candidates blah blah blah.”

No recruiter called. Just an automated rejection.

And after seeing this happen enough times, it’s easy to start questioning yourself.

Maybe I’m not qualified. Maybe my experience isn’t relevant. Maybe the market is just impossible right now. But before you assume you’re the problem, it’s important to understand what has changed.

Many companies are now receiving hundreds, and sometimes thousands, of applications for a single role. Recruiters simply do not have the capacity to manually review every resume that comes in.

As a result, hiring teams increasingly rely on applicant tracking systems, automated screening workflows, ranking algorithms, and AI-assisted recruiting tools to help narrow the field.

The result?

Many candidates are being filtered out before a recruiter ever reviews their resume.

That’s why you’re seeing rejection emails arrive at strange hours.

That’s why you’re hearing stories of people applying to 200, 500, or even 1,000 jobs with very little response.

And that’s why old job-search advice is becoming less effective.

Today, there are only two strategies that consistently improve your odds of getting noticed.

Strategy #1: Apply Within the First 24 Hours

This is the first most important recommendation in this entire article.

If a role has been open for several days and already has hundreds of applicants, your chances drop dramatically.

Recruiters typically start reviewing candidates as applications arrive. By the time a posting has been live for a week, many already have a strong shortlist.

The best opportunity exists when a role is newly posted.

In fact, if I were job searching today, I would prioritize speed over volume.

I would rather be one of the first 25 applicants to a highly relevant role than applicant number 1,500 to a role that’s been open for two weeks.

Apply to relevant jobs within the first 24 hours of posting.

That is often your highest-probability path to getting seen.

Share

Strategy #2: Optimize for Ranking, Not ATS

Most job seekers ask the wrong question.

They ask:

“Will my resume pass the ATS?”

That’s not the right question anymore.

A resume can be perfectly readable by an ATS and still perform poorly.

Why?

Because ranking matters.

Recruiters are often reviewing the strongest candidates first. If your experience isn’t clearly positioned, categorized, and aligned with the role, you may never make it to the top of the pile.

This is where most candidates struggle.

They submit the same generic resume to every job.

Meanwhile, the strongest candidates are tailoring their positioning, surfacing the most relevant experience, matching recruiter search behavior, and making it obvious why they’re a fit.

That’s why I developed a six-step Resume Ranking Workflow.

The workflow isn’t about keyword stuffing.

It’s about understanding who gets hired, how recruiters search, how you compare against the competition, and how to position yourself so that your experience is immediately recognizable.

In the next section, I’ll walk through the exact six-step process I use to evaluate, position, and optimize a resume for today’s AI-assisted hiring environment.

Below I have mentioned exact prompts that will get your resume to rank higher, along with applying to jobs that have been posted for less than 24hours.

Step 1: Understand Who Gets Hired

ChatGPT PROMPT below

You are a recruiter hiring for this role.
You have 1,500 applicants.
You need to narrow the pool to 30 recruiter screens.
Analyze this job description.

Tell me:

1. What candidate archetype would make the shortlist?
2. What experience patterns appear repeatedly among likely interview candidates?
3. What resume signals would immediately stand out?
4. What resume signals would cause rejection?
5. What keywords are actually important?
6. Which keywords are irrelevant even though they appear in the JD?
7. What would the top 5 resumes have in common?
8. If you had 10 seconds to scan a resume, what would you look for?

Job Description:
[paste]

Read more

Before we get into it, I want you to sit with one number: 53%.

That’s the share of job seekers who got ghosted by an employer in the past year, according to a report from Criteria covered by Fortune this spring. That’s a three-year high, up from 38% in 2024. If you’ve been sending applications into the void and hearing nothing, you’re not imagining the void. It’s real, it’s measured, and it’s growing.

But here’s the part nobody tells you, and it’s the reason I’m writing today’s issue.

You probably think of every application as an independent shot. Fifty applications, fifty separate chances, fifty different humans who might see something in you. That assumption just collapsed.

A team of Stanford researchers (Bommasani, Bana, Creel, Jurafsky, and Liang) recently published the largest study of algorithmic hiring ever conducted: 3.4 million real job applicants, 4 million applications, 156 employers across 11 industries, with every single application scored by AI models from the same vendor. They call the phenomenon they studied “algorithmic monoculture”: most companies don’t build their own screening AI, they buy it from the same handful of vendors. Which means the “no” you got from Company A and the “no” you got from Company B may not be two opinions. They may be one opinion, duplicated.

few things surfaced:

• Systemic rejection is real. Among candidates who submitted four or more applications, roughly 1 in 10 were rejected across the board by the shared algorithm, far more than you’d expect if companies were making independent decisions. The researchers had theorized for years that shared AI vendors could lock some people out of interviews entirely. This was the first time they could prove it in real hiring data.

• The bias is measurable. The study found clear racial disparities: about 26% of applications from Black applicants and 15% from Asian applicants were routed to positions where the algorithm adversely impacted them.

• The audits missed it. Here’s the kicker: the vendor had published audits showing no measurable bias. The disparities only appeared when researchers analyzed positions one by one instead of in aggregate. Even the safeguards are a black box.

And the scale of all this is no longer niche. As of early 2026, roughly 87% of companies use AI somewhere in recruitment, and 99% of the Fortune 500 have integrated it into hiring. AI-conducted interviews have more than tripled in two years (10% → 34%). And in one compilation of employer surveys, only 29% of companies keep a human reviewing every AI rejection; about 1 in 5 let the AI reject candidates at any stage with no human review at all.

Share

I lived this last year. When I was job hunting, my resume wasn’t even passing the ATS in many, many instances. I kept updating it, and updating it, and updating it again until, I guess, I finally cracked whatever invisible threshold the system wanted. I had no idea what changed or why. That experience is exactly what sent me down this rabbit hole, and what I found is the topic of today’s newsletter: the machine reading your resume has fundamentally changed, and most candidates are still playing by the old rules.

The keyword-stuffing era is over. Welcome to semantic matching.

🧠 From Keywords to “Skills DNA”: What Changed

Previously: Your resume needed the exact phrase “stakeholder management” to match a job posting asking for “stakeholder management.” The game was crude, but at least it was transparent.

In 2026: Modern screening platforms use semantic matching: they evaluate meaning, not just words. Companies like Eightfold(dot)ai have built deep-learning models that infer skills and career trajectories from unstructured resume data and match candidates on capability, not keyword alignment. Industry analysts now describe semantic, skills-based matching as operationally necessary for any team handling 250+ applications per role, which is to say, every team you’re applying to. And note who’s most exposed: platforms like Eightfold primarily serve large enterprises (5,000+ employees). If you’re applying to big tech, this is what’s reading you.

In practice: if you wrote “aligned cross-functional teams across engineering, design, and marketing to deliver product launches,” the system infers you have stakeholder management, cross-functional leadership, and project coordination, even without those exact phrases.

A former teacher who “designed curriculum for 150+ students and presented to parent committees” can now match to a corporate training role, because the AI recognizes the underlying Skills DNA: instructional design, public speaking, stakeholder communication.

In some ways, this is genuinely good. You’re no longer competing on who stuffed the most keywords. You’re being evaluated on the substance of what you’ve actually done.

But let me play devil’s advocate, because this is where it gets uncomfortable.

You have no idea what the AI is inferring from your resume, or what “Skills DNA” profile it’s building of you. When keyword matching ruled, the rules were visible: job posting says “stakeholder management”? Put “stakeholder management” in your resume. Done.

Now? The AI might extract “strategic thinking” from one of your bullets, or it might not. You’re writing blind, hoping a black box interprets your experience the way you intended, with no feedback loop telling you what skills it extracted or why you were rejected

Share

And the Stanford study shows why this matters so much: when the black box gets you wrong, it may get you wrong everywhere it’s deployed. Same model, same inference, same rejection, repeated across every employer that licensed it. Regulators have noticed, by the way: starting August 2026, the EU AI Act classifies hiring AI as “high-risk,” requiring bias testing, human oversight, and transparency, with serious fines attached. The US has no equivalent yet. For now, you’re on your own.

So the practical question becomes: how do you write a resume for a machine that reads meaning?

⚙️ The Three Stages of Screening Your Application

Quickly, because you need this context for what follows:

1. Parsing. The ATS extracts your employers, titles, skills, and education into structured data. Bad formatting kills you here, before any AI even evaluates you.

2. Semantic matching. Transformer-based models (the same family of technology behind ChatGPT) score the similarity between your experience and the role. This is where your Skills DNA gets built.

3. Human review. Most companies still have a recruiter look at top-scored candidates. Your entire goal is to survive stages 1 and 2 so a real person finally sees your work.

Subscribe now

🎯 Rewriting Your PM Experience for Semantic Matching

Each example follows Context → Action → Result:

Example 1: Feature Launch

❌ Keyword-stuffed (old way): “Responsible for product roadmap and feature prioritization. Managed stakeholders and delivered features.”

✅ Semantic-friendly (new way): “Defined Q3 product roadmap by analyzing user feedback from 2,000+ customers and competitive intelligence, prioritizing 8 features that reduced churn by 23%. Facilitated weekly sprint planning with 12-person engineering team and bi-weekly executive updates to align on trade-offs.”

What the AI extracts: Product strategy, data-driven prioritization, stakeholder management, churn reduction, cross-functional leadership, communication skills, agile methodology.

Example 2: Discovery & Research

❌ Keyword-stuffed (old way): “Conducted user research and customer interviews. Gathered requirements.”

✅ Semantic-friendly (new way): “Led discovery for payment redesign by conducting 25 user interviews and analyzing 6 months of checkout funnel data (50K+ sessions). Identified 3 critical friction points causing 40% drop-off. Created journey maps and wireframes with design team, validated with A/B test reaching 15K users.”

What the AI extracts: User research, qualitative analysis, quantitative analysis, problem identification, data interpretation, A/B testing, UX collaboration, hypothesis validation.

The pattern: each bullet shows the situation (customer problem, business need, team size), what you specifically did (methods, tools, collaboration), and what changed (metrics, outcomes). Semantic matching recognizes this structure and extracts the underlying skills automatically.

💼 What You Should Actually Do

1. Rewrite your top achievements using Context → Action → Result.

Bad: “Managed product roadmap” Good: “Prioritized 12-month roadmap based on analysis of customer requests (3,000+ tickets), revenue potential, and technical feasibility, resulting in 40% faster feature delivery”

2. Make your resume machine-readable. Standard headers (”Experience,” “Skills,” “Education”). No graphics, tables, or multi-column layouts. Clean, consistent fonts. Save as .docx or text-based PDF, never an image-based PDF. Test with a free ATS checker like Jobscan or Resume Worded. Remember: stage 1 parsing failures end your candidacy before any intelligence, artificial or human, ever weighs in.

3. Emphasize transferable skills in context. Non-technical roles live and die on demonstrated adaptability:

• Cross-functional collaboration: “Coordinated 8-person team across eng, design, data”

• Data-informed decisions: “Analyzed user behavior data from 10K+ sessions”

• Ambiguity navigation: “Launched MVP with 60% of planned features based on customer urgency”

• Technical communication: “Translated customer needs into technical requirements for API team”

These are the Skills DNA markers semantic matching extracts, and exactly what hiring managers want anyway.

4. Quantify everything. Team size, revenue, users, retention, time saved. Numbers are how the model understands scale and significance. “Responsible for product strategy” tells the AI nothing. “Developed 3-year product strategy based on TAM analysis ($50M opportunity) and customer interviews (40+ B2B buyers), securing $2M budget approval” tells it everything.

5. Don’t keyword-stuff: it now works against you. A skills section reading “agile, scrum, kanban, jira, confluence, roadmapping, stakeholder management...” is detectable gaming. The models look for skills demonstrated in context, not listed in a row.

6. Keep your digital footprint for the humans, not the machines. One honest correction to the standard advice here: most screening AI is not crawling your Medium posts or LinkedIn activity. Your portfolio, case studies, and public writing matter at two other points: when recruiters source candidates proactively, and at stage 3, when a human finally googles you. Build it for them. In 2026, your resume gets you screened; your work gets you hired.

⚠️ The Part No Resume Rewrite Fixes

I want to be honest with you about something most career content won’t say: you can do everything above perfectly and still get systemically rejected.

That’s the real lesson of the Stanford study. If a shared model scores your profile poorly (fairly or not), that score follows you across every employer using it, and there is no appeals process and no feedback loop. Some portion of your silence is structural, not personal. I’m telling you this for two reasons.

First, because it should take some weight off your shoulders. The endless cycle of “what’s wrong with my resume?” assumes the system is giving you an accurate signal. It isn’t, always.

Second, because it changes strategy. If the algorithmic front door is partially broken, the rational move is to not to rely on it exclusively. Sourced candidates, meaning people a recruiter, hiring manager, or employee knew about before the application, get hired at roughly 8x the rate of cold applicants. Optimizing your resume for semantic matching is necessary. It is not sufficient. (The “how” of going around the front door deserves its own issue. That’s coming in a future newsletter.)

So: semantic matching evaluates substance over buzzwords, which feels fairer, but it does so invisibly, at massive scale, with documented bias its own audits failed to catch, increasingly with no human in the loop.

Is this progress or just a more sophisticated black box?

Tell me in the comments, and tell me your ghosting stories. I read every one.

That’s all for today!

Nazuk

Subscribe now

Sources referenced in this issue

• Bommasani, Bana, Creel, Jurafsky & Liang, “Algorithmic Monocultures in Hiring,” Stanford Digital Economy Lab: https://digitaleconomy.stanford.edu/publication/algorithmic-monocultures-in-hiring/

• Stanford Digital Economy Lab Q&A with the study authors: https://digitaleconomy.stanford.edu/news/qa-algorithmic-monoculture/

• HR Dive coverage of the study: https://www.hrdive.com/news/how-a-hiring-algorithm-is-audited-can-disguise-bias-study-finds/821852/

• Fortune coverage of the Criteria ghosting report (53% of job seekers ghosted)

• CoverSentry, “AI in Hiring Statistics 2026” (human oversight and AI interview data): https://www.coversentry.com/hiring-ai-statistics

• Limelight Digital, AI recruitment adoption statistics (87% of companies, 99% of Fortune 500): https://www.limelightdigital.co.uk/ai-recruitment-statistics/

• SQ Magazine, AI recruitment statistics (EU AI Act enforcement timeline): https://sqmagazine.co.uk/ai-recruitment-statistics/

• HackerEarth & Talentpilot analyses of semantic screening tools incl. Eightfold AI

• Careery, job search conversion data (sourced vs. cold applicant hire rates): https://careery.pro/blog/job-search-strategy/why-is-it-so-hard-to-find-a-job

Then you open LinkedIn and see people somehow landing roles at OpenAI, Anthropic, Cursor, Perplexity, Meta, Google, and AI startups paying $250K–$500K+ total compensation.

So what is actually happening?

Last year, after my maternity leave, I landed an AI Product Ops role with Meta. One of the greatest experiences of my life. and THE ONE that has allowed me to use that experience as “proof-of work” for AI experience.

A huge challenge for every PM is; it seems “every” job needs AI experince but to get that AI experience you need the job in the first place.

My thoughts are plain and simple at the moment on how to think about thsi chiken and egg situation (simply to unlock the complexity in my mind and somewhat reduce my own anxiety around it), i.e., the hiring bar is changing with this new technological shift so we need to adapt.

Most mid-career PMs are still preparing for the old version of product management. But Mid-Career PMs will also be the most advantageous in this situation, because of all the skills they have honed over the years (my2cents)

I hope this newsletter today will provide you with some guidance and direction to take the next step.

Today’s issue is about one important thing: understanding the market before you start applying.

Because most people skip this step entirely.

Share

The Fear Is Real — But So Is The Opportunity

A lot of experienced PMs are feeling stuck right now.

Especially people with 7–15 years of experience, strong execution backgrounds, traditional SaaS or product experience, and roadmap or stakeholder-heavy roles.

So many of you have shared your fear, and I have been asked: “Do I need to become an engineer now?” “Is prompt engineering enough?” “Am I already behind?” “How do I even transition into AI?”

I want to share something that might help.

A PM I know — 11 years of experience in fintech, no engineering background — felt exactly this way six months ago. She spent 30 days studying AI PM job descriptions, built one practical proof-of-work project (an AI-assisted customer support workflow redesign), and rewrote her LinkedIn around systems thinking and automation. She landed a Senior AI PM role at a Series B startup at nearly double her previous total comp.

She was not a technical. She layered AI capability on top of existing domain expertise. That is the real playbook.

The problem for most experienced PMs is not that they are unqualified. It is that the market shifted faster than most people realized.

Traditional PM hiring is optimized for roadmap management, coordination, communication, delivery, and stakeholder alignment.

AI PM hiring increasingly evaluates systems thinking, workflow design, ambiguity handling, rapid prototyping, AI fluency, product intuition, technical judgment, and speed of execution.

That does not mean your previous experience became irrelevant. It means you now need to layer AI capability on top of your existing domain expertise. That is a very different strategy from starting over.

First: Is The Market Actually Real?

Yes. Very real.

Glassdoor currently shows more than 1,300+ AI Product Manager roles in the U.S.

Salary estimates show:

• U.S. AI PM average total compensation around ~$196K

• California AI PM compensation averaging ~$220K

• Top percentile compensation exceeding ~$320K+

(U.S. AI PM salaries | California AI PM compensation)

Levels.fyi data also shows senior product compensation in AI-heavy markets like Bay Area, New York approaching or exceeding $300K total compensation.

One important thing to understand about these numbers: a significant portion of $300K+ packages — especially at startups — is equity. A $300K offer at an early-stage AI startup might be $160K base + $140K in equity. That equity could be worth far more, or far less, depending on how the company performs.

With FAANGs and AI-Natove companies like Open-AI, Anthropic, Perplexity, comps can be upward of 500K+ to 700k+.

But the opportunity is absolutely real. The bigger issue is that most PMs are preparing for the wrong thing.

Share Product Career

What Companies Are Actually Hiring For

Last week I shared a post where I looked at Non-FAANGs. This week I spent time reviewing AI PM job descriptions across OpenAI, Anthropic, enterprise AI companies, healthcare AI companies, and workflow automation startups.

There are very clear patterns.

Pattern 1: Companies want AI workflow thinkers, and here’s why

This is one of the biggest shifts happening right now. Companies are not just looking for PMs who can “use ChatGPT.” They want PMs who can redesign entire workflows using AI.

The reason: most enterprise AI ROI does not come from chat interfaces or AI features bolted onto existing products. It comes from operations — automating the work that slows companies down, reducing the cost of decisions, and redesigning workflows that were built for a pre-AI world.

OpenAI’s Enterprise PM role explicitly references AI-enabled workflows, automation systems, agentic workflows, and enterprise productivity transformation.

This matters because many PMs still think AI PM work is mostly prompting, chat interfaces, or AI features. But increasingly, the value is operational.

Applied AI is a workflow redesign.

Subscribe now

Pattern 2: AI fluency matters more than “AI expert” credentials and here’s why

Continue reading by becoming a paid subscriber.

You will get access to the following:

• Pattern 3:

• AI PM Archetypes you need to understand

• Generic ChatGPT Traps

• How to Talk About AI in Interviews (Without Sounding Like a Buzzword Generator)

• NEWer questions that are now asked in interviews

• Red Flags That Will Get You Filtered Out

• Self assessment

• The Fastest Path Into AI PM

• Resources
  

Read more

How PMs Are Quietly Becoming Product Builders

There’s a shift happening to the PM role right now, and most product managers are hearing about it but have no clue where to begin and how to even think about this shift, given the noise on LinkedIn.

The companies moving fastest (Your standard FAANG stack and More) have quietly stopped looking for PMs who just write PRDs, coordinate stakeholders, manage roadmaps, and run meetings. Those skills still matter. But they’re no longer the thing that sets you apart.

AI changed the leverage equation completely. And the most valuable PMs in some of these top companies I know are responding to it by quietly becoming product builders.

Not engineers. Builders. There’s a difference, and it matters more than ever — so let me explain what I mean.

Share

The Old PM Playbook Worked for a Reason

For years, the PM role was built around coordination. A strong PM gathered requirements, aligned stakeholders, prioritized the roadmap, managed timelines, wrote the docs, and kept communication flowing between teams.

And honestly? That made complete sense.

Engineering execution used to be the bottleneck inside almost every organization. Building software was expensive, slow, and resource-constrained. So PMs became the information layer in waterfall and Agile shops.

They translated business goals into engineering requirements, cut down communication overhead, and kept everyone pointed in the same direction.

That coordination created enormous value. I’m not here to tell you it didn’t. But the bottleneck has moved. And when the bottleneck moves, the job changes with it.

AI Changed Where the Bottleneck Lives

Today, an idea can become something tangible in an afternoon.

A PM can now prototype a workflow, generate wireframes, build a lightweight MVP with no-code tools, spin up an AI-generated landing page to test messaging, automate the boring analysis, simulate a user flow, or stand up an internal tool — all before a single engineer is pulled in.

That’s not a small upgrade. That’s a different game.

And that one shift quietly rewrites what companies value in a PM. The people creating the most leverage aren’t the ones writing the cleanest specs. They’re the ones killing bad ideas early and building conviction around good ones — fast, and with evidence. AI is what makes that speed possible.

Share

The Rise of the Product Builder

I think we’re entering the era of the Product Builder PM.

A Product Builder prototypes ideas early, validates assumptions quickly, makes concepts tangible, experiments rapidly, and drives conviction through evidence instead of opinion. They use AI to create momentum where there used to be a waiting period.

Let me be clear about what this is not.

It does not mean you suddenly have to become a full-time software engineer. That’s not the point, and chasing it will waste your time.

The point is that the distance between idea and execution has collapsed. The PMs who can operate inside that new reality — turning a hunch into something people can actually click on — are going to be worth far more than the ones who can only describe what should exist.

The future PM isn’t just a spec writer. They’re an orchestrator of systems, workflows, experiments, and AI leverage.

What I’m Seeing Up Close

Over the past year, a few patterns have become impossible to ignore.

1. AI-first workflows became normal shockingly fast. Teams stopped asking “Should we use AI here?” and started asking “How do we redesign this entire workflow around AI?” That’s a completely different question, and it happened in months, not years.

2. Experimentation cycles compressed. Faster prototypes, faster validation, faster learning. The expectation now is that you show up with something, not just a slide about something.

3. Initiative became the differentiator. The PMs standing out aren’t waiting for permission to explore. They’re building quick demos, testing concepts on their own, and using AI to create momentum before formal prioritization even begins.

If you don’t believe me, go read the job descriptions yourself. The shift is written into them in plain English.

Don’t Take My Word For It — Read the JDs

I went and pulled real, currently-live PM job postings from some of the most respected companies in tech. and FAANGs are at the forefront of this shift, so I looked at some non-FAANGs that are still respected in the space.

Intuit — Lead Product Manager. The JD literally instructs candidates to Leverage AI to analyse and experiment aka vibe code.

Capital One — Across multiple PM postings, “Pioneer AI QA and Feedback Loops”

Stripe — General PM listings describe a process centered on iteration, where you “create prototypes, get them in front of customers to test, and repeat.” Stripe even has a dedicated AI Skills needed section in the “minimum requirements”.

Chime — The Lead PM, Hands-on AI experience.

All of these jobs in SF range from 150K-250K base salaries.

Notice what’s missing from all of these JDs: a list of frameworks. A demand for spec-writing fluency. A required number of years running standups. That stuff still matters, but it’s no longer the headline.

The headline is: build, vibe code, hands on AI work-ex.

And here’s the part that should get your attention — these are some of the highest-paid PM roles on the market. According to job-market analyses, AI PMs at companies like OpenAI, Anthropic, and Google are commanding compensation 30–40% above traditional PM benchmarks, with total comp at the senior level frequently crossing $400-600K. The market is putting real money behind this redefinition. Pretending it’s not happening is an expensive position to hold.

I’m also watching PM interviews change in real time.

More companies are asking candidates to critique a product live, prototype an idea on the spot, demonstrate systems thinking, walk through an AI-assisted workflow, or explain how they’d validate something cheaply. either using an AI tool.

Several companies have now replaced their Product Sense or technical round with a VIbe coding round. (will write more on this next week)

I don’t think this slows down. If anything, it accelerates.

For example, this is a REAL interview question I was asked in a Live Vibe Coding interview:

(Redacted company name because of NDA)

Leave a comment

“But Doesn’t This Just Make PMs Redundant?”

I get this question a lot. Here’s how I see it.

No, but it does make one kind of PM redundant: the one whose entire value was being the translation layer between business and engineering. When ideas were expensive to test, that role was essential. Now that they’re cheap to test, pure coordination is the part AI and faster tooling absorb first.

The judgment doesn’t go away. Knowing what to build, why, and for whom gets more valuable, not less — because now you can find out faster, and the cost of being wrong drops. Building isn’t replacing product thinking. It’s giving product thinking a faster feedback loop.

Now I truly think that “knowing what to build” is a Skill of CEOs.

It’s not an easy skill to build.

Imagine how company CEOs bet on what to build next, and how they strategize. PMs somewhere hone that skill their entire career, if they do well and if they get the opportunity to do so.

The risk isn’t that you become obsolete. The risk is that you keep optimizing for the old version of the job while the new one is already being hired for.

How to Start

Here’s the good news: you don’t have to master all of this overnight. You just have to start building leverage, gradually and consistently. Here’s where I’d begin.

1. Learn one prototyping workflow — just one (don’t let any creator, including myself, tell you the hottest tool to learn, really) instead:

Pick a single lightweight stack and run it until it’s muscle memory. Claude + Notion. ChatGPT + Figma. Cursor + v0. Replit + AI agents. The goal isn’t a polished product; it’s collapsing the friction between idea and thing I can show someone. Start tiny: redesign an onboarding flow, mock a dashboard, fake an MVP for a product idea you’ve been sitting on.

2. Build tiny internal tools.

The fastest way to become AI-native is solving small, real workflow problems. Automate your meeting summaries. Build a customer-feedback analyzer. Draft a prioritization assistant. Generate first-pass PRDs. These little projects teach you prompting, automation logic, edge cases, and systems thinking far faster than any course — and they build the confidence to go bigger.

3. Validate before you plan.

Most PMs are still over-investing in planning before they’ve earned the right to plan. Flip it. Before the full PRD, ship a landing page, run a fake-door test, prototype the experience, or interview users against an AI-generated concept. Optimize for learning velocity. The faster you reduce uncertainty, the more valuable you are.

4. Treat prompting as structured product thinking not other way around.

Prompting isn’t magic, and it isn’t a party trick. It’s the same skill as writing a good brief: provide context, define constraints, structure the problem, communicate the outcome you want, and iterate. If you can write a clear PRD, you can write a clear prompt. They’re the same muscle. But the internet has us so confused.

5. Think in workflows, not features.

AI-native PMs zoom out. What process is actually broken? Where’s the friction? What’s repetitive enough to automate? Where do humans still need to stay in the loop? Features are the old unit of thinking. Workflows are the new ones.

Subscribe now

Resources Worth Your Time

A short, opinionated list — the goal is to build small things consistently, not to consume endlessly.

Tools: Claude, ChatGPT, Perplexity, Cursor, Replit, v0 by Vercel , Notion AI, Figma AI

Learning: DeepLearning.AI short courses, OpenAI Cookbook, Anthropic prompting guides, Lenny’s Newsletter, a16z AI

People to follow: Lenny Rachitsky, Ethan Mollick, Andrej Karpathy , Claire Vo, Shreyas Doshi, Wes Kao

5 free Anthropic courses:

1. Claude 101

2. AI Fluency: Framework & Foundations — the strategic mental model

3. Introduction to Claude Cowork

4. Claude Code 101

5. Building with the Claude API — for the more technical PMs

Projects to try this month: an AI onboarding assistant, a customer-feedback summarizer, a PRD copilot, a feature-recommendation prototype, an automated roadmap summary

The PM role is evolving; quietly, and faster than most people are reacting to.

The distance between idea and execution just collapsed. The only question left is whether you move into that space before someone else does.

The people who adapt early will have outsized leverage over the next few years. Not because they collected more tools. Because they learned how to turn ideas into momentum while everyone else was still writing about them.

The biggest mistake you can make right now is preparing for the old version of the job while the new one is already being hired for.

So here’s my one ask: don’t read this newsletter, nod and move on.

Pick one thing from the list above and do it this week. Prototype something rough. Automate one annoying task. Ship a fake door. Then hit reply and tell me what you built — I read every response, and I’ll feature the best ones in a future issue, and on my socials (who knows what kinda of visibility that can bring for you, if that’s something you are looking for.)

If you found today’s newsletter useful, share it with a friend and leave me a comment. I’ll see you next week.

-Nazuk

Subscribe now

As I was wrapping up my maternity leave, I was in this unique situation where my husband and I were unemployed at the same time (his company had lay-offs and he had been with them for decade+ for context).

So to minimize my time to land a job, I got strategic. Behavioral rounds are your foot in the door for any interview. They dictate your first impression/ eligibility/ capabilities, critical thinking/ potential etc for the role.

…not something you can take lightly. So after applying the tactics I shared below, I am happy to report both my husband and I landed roles in 3 months, start to finish, in the slowest hiring time of the year, Oct/Nov.

I decided to write a FULL Playbook on Behavioral Interviews - Inside the interviewer’s Head

Below are the 6 Moves I have shared in the playbook. If you like these You will find the playbook even more valuable. Link at the bottom.

Here are 6 tactics you can employ

These are specific moves designed to work with the psychology we just covered because there’s no point learning tactics that ignore how interviewers actually think.

1/

Front-Load Your Strongest Story

Don’t save your best for last. Primacy bias means your first behavioral answer sets the anchor for the whole interview. Choose a story that’s high-stakes, clearly structured, and shows judgment, especially how you think through decisions.

If the interviewer’s first note is “sharp thinker, strong product instinct,” everything after that gets a tailwind. If their first note is “seemed nervous, rambled a bit,” you’re now fighting uphill for the rest of the hour.

Here’s how to pick your opener: Ask yourself: which story makes me sound like the person I want to be in this role? Pick the one that shows how you think, even if it wasn’t your biggest project.

2/

Match the Interviewer’s Mental Model

Every interviewer has an unconscious picture of what “great” looks like. That picture is shaped by their own career. A former founder lights up at scrappiness and speed. A senior PM at a big company values cross-functional influence and process rigor.

Move: check their LinkedIn before the call. Five minutes of research tells you what they value. Their background is your cheat code.

Share

3/

Use the “So What” Close

Most candidates end their answer at the Result: “We shipped the feature.” “Retention improved by 15%.” That’s fine, but it leaves the interpretation to the interviewer.

Great candidates add one sentence that frames why the result mattered: “That feature became the second-largest driver of new user activation that quarter, which directly informed our Series B narrative.” Now the interviewer doesn’t just hear a result; they hear business impact and strategic awareness.

4/

Name the Tradeoff

PMs are hired to make decisions under uncertainty. When you tell a story, explicitly call out the tradeoff you faced and why you chose the path you chose.

“We could have done X, which would have been faster, but I chose Y because...” That sentence signals judgment more than any polished success story. It shows you understand that every decision has a cost, and you’re comfortable making the call and owning the outcome.

5/

Manage the Portfolio, Not Just Each Story

Four behavioral questions aren’t four separate answers; they’re a portfolio. The hiring manager is unconsciously evaluating coverage: did this person show range, or do they only have one gear?

The move: Before the interview, map your stories to competency categories (leadership, data/analytics, conflict resolution, ambiguity, customer focus, technical collaboration). Make sure your prepared set covers at least 4 different dimensions.

6/

Pause Before You Answer

This one feels counterintuitive, but a 3-second pause after the question is asked signals confidence: it says “I’m choosing the right story,” not “I’m blanking.” Interviewers notice when someone takes a beat to be deliberate.

The candidates who jump in immediately often ramble because they started talking before they figured out where they were going. The ones who pause, then deliver a structured answer, come across as clear thinkers. Which is what the interview is actually testing.

Subscribe now

That’s all for today.

-Nazuk

👉 Here is the Link to the full guide.

QUICK NOTE:

You will need to create your login (one time only and FREE), but that allows you to access all of the prep material I share (I will be adding more resources next week).

I finally have one place for everyone to access Resources. Substack doenst unfortenaly have a system I can provide you all with additional resources.

Today’s newsletter is for the other 99% who tried this method and haven’t gotten a call back.

without any context, you’re essentially asking a stranger who’s never met you to convince an employer why you’re the perfect hire.

with this kind of prompting you get a document full of generic corporate speak that says absolutely nothing about you.

When you feed AI minimal information, it fills the gaps with what it knows best: patterns. And those patterns are built from millions of resumes it’s been trained on. So your “unique” resume ends up sounding like this:

• “Results-driven professional with proven track record...”

• “Excellent communication skills and team player...”

• “Passionate about driving organizational success...”

Hiring managers see hundreds of these. They’ve developed pattern blindness to this language. Your resume becomes wallpaper.

Why Context Is Everything

AI isn’t psychic. It doesn’t know:

• That you rebuilt the entire customer onboarding process and reduced churn by 34%

• That you taught yourself Python to automate reporting that saved your team 15 hours weekly

• That you managed a product launch during a company restructuring with zero budget increase

• The specific technologies, methodologies, and frameworks you’ve mastered

• How you actually communicate, solve problems, or lead teams

~you get where I am going with this~

Subscribe now

Without the details, AI invents plausible-sounding but ultimately meaningless filler.

The Right Way to Use AI for Your Resume

Think of AI as an exceptionally skilled writer and editor, but one that needs your raw material to work with.

Feed it your career assets:

• Performance reviews and feedback

• Project descriptions and outcomes with metrics

• Awards, recognitions, and specific achievements

• Your actual job descriptions and responsibilities

• Problems you solved and how you solved them

• Technologies and tools you’ve used extensively

• Your authentic communication style from emails or reports

Then give it clear instructions like:

• “Here are five projects I led. Transform these into achievement-focused bullet points that quantify impact.”

• “I’m applying for a senior product role at a healthtech startup. Reframe my experience to emphasize relevant skills.”

• “Make this sound more confident and senior-level while keeping my voice.”

The Difference This Makes

A context-rich prompt produces a resume that:

• Uses specific, verifiable achievements instead of vague claims

• Speaks in your actual voice and experience level

• Aligns precisely with roles you’re targeting

• Tells a coherent story about your career progression

• Stands out because it contains real substance

Meanwhile, the generic “write me a resume” approach produces something that could describe thousands of other candidates. It might look professional, but it’s optimized for nothing except looking like a resume.

Your Resume Is a Sales Document

Would you let a salesperson pitch your most important product without giving them any information about it? That’s exactly what you’re doing when you ask AI to write your resume from scratch.

Your career deserves better than AI’s best guess. Give it the data, and you’ll get back something worth shortlisting.

Share

Stop asking AI to create. Start asking it to elevate what you’ve already built.

I promise the unique you will shine.

That’s all for today

-Nazuk

P.s.

Bonus: I wrote about how how you can build a light-weight RAG system without engineering that went mildly viral. Read here if you missed.

PLUS

Pair that with my Newly launched interview dashboard that has AI prompts to help you practice with AI on your own time in a meaningful way with YOUR career stories.

Buy Now

I promise within 3 weeks you will be ready for your first interview if you put in the effort.

Here is another one of the essay you might find useful while preparing your career stories unique storytelling ways

As we head into the weekend, I’ve got some crazy intel to share with you all this week on AI recruiting. I recently discovered how keywords and stuffing keywords in your resume is so dead.

which explained a lot of why last year when I was job hunting my resume did not even pass ATS in many many instances until I kept updating it updating it and updated to a point where I guess I did pass the ATS but I’ve learned how there are companies now out there (one of the examples is Eightfold.ai) who have started building the back-end AI recruiting intelligence that builds on semantic matching.

Which is the topic of my newsletter today.

And also know that this newsletter is the first you are hearing about it because this is going to become more and more prevalent in hiring processes in all of 2026.

Our generation is living in the most exciting and anxiety inducing times in the history of job search.

Today I want to tackle the question of

• How do you get past the ATS (Applicant tracking system)?

• Is mass application a good strategy?

• The quality vs quantity?

• What is skill inference vs. semantic matching and how this has created the doom loop

Share

If you’re a Product Manager or any non-technical professional in tech, the job market has fundamentally changed and most candidates don’t realize they’re playing by outdated rules.

Here’s what’s actually happening: 79% of organizations have integrated AI into their hiring systems, up from 43% two years ago. But here’s the twist: while technical roles have adapted, non-technical professionals are often caught off-guard by how differently their applications are now being evaluated.

The keyword-stuffing era is over. Welcome to semantic matching.

🧠 From Keywords to “Skills DNA”: What Changed

Previously: Your resume needed the exact phrase “stakeholder management” to match a job posting asking for “stakeholder management.”

If you wrote “aligned cross-functional teams across engineering, design, and marketing to deliver product launches,” the system now infers you have stakeholder management, cross-functional leadership, and project coordination skills—even without those exact phrases.

A former teacher who “designed curriculum for 150+ students and presented to parent committees” can now match to a corporate training role because AI recognizes the underlying Skills DNA: instructional design, public speaking, and stakeholder communication.

In some ways I think this is good because You’re no longer competing on who stuffed the most keywords. You’re being evaluated on the substance of what you’ve actually done. The systems look for demonstrated impact, not buzzword bingo.

But If I played devil’s advocate

While semantic matching sounds fairer, there’s a darker reality: you have no idea what the AI is actually inferring from your resume. or what type “DNA” they are building of your Skills. You have no visibility.

When keyword matching ruled, the game was transparent. Job posting says “stakeholder management”? Put “stakeholder management” in your resume. You knew the rules.

Now? The AI might infer you have “strategic thinking” from one phrase, or it might not. You’re writing blind, hoping the black box interprets your experience the way you intended. There’s no feedback loop telling you what skills the system extracted or why you were rejected.

Share

⚙️ What’s Really Screening Your Application

Your application goes through three stages:

Stage 1: Parsing & Normalization (Immediate)
The ATS extracts your employers, titles, skills, and education into structured data. Poor formatting kills your chances here before AI even evaluates you.

Stage 2: Semantic Matching (Automated)
AI models (using transformer-based technology similar to ChatGPT) analyze the semantic similarity between your experience and job requirements. This is where your “Skills DNA” gets evaluated.

Stage 3: Human Review (If you pass)
88% of companies acknowledge their automated systems reject qualified candidates—so most still have human review for top candidates. Your goal: get past Stage 2 so a real person sees your work.

🎯 If you were to rewrite some of Your PM Experience

Here’s how to write for semantic matching.

Each example follows Context → Action → Result:

Example 1: Feature Launch

❌ Keyword-Stuffed :
“Responsible for product roadmap and feature prioritization. Managed stakeholders and delivered features.”

✅ Semantic-Friendly :
“Defined Q3 product roadmap by analyzing user feedback from 2,000+ customers and competitive intelligence, prioritizing 8 features that reduced churn by 23%. Facilitated weekly sprint planning with 12-person engineering team and bi-weekly executive updates to align on trade-offs.”

What AI extracts: Product strategy, data-driven prioritization, stakeholder management, churn reduction, cross-functional leadership, communication skills, agile methodology

Example 2: Discovery & Research

❌ Keyword-Stuffed (Old Way):
“Conducted user research and customer interviews. Gathered requirements.”

✅ Semantic-Friendly (New Way):
“Led discovery for payment redesign by conducting 25 user interviews and analyzing 6 months of checkout funnel data (50K+ sessions). Identified 3 critical friction points causing 40% drop-off. Created journey maps and wireframes with design team, validated with A/B test reaching 15K users.”

What AI extracts: User research, qualitative analysis, quantitative analysis, problem identification, data interpretation, A/B testing, UX collaboration, hypothesis validation

The Pattern: Each bullet shows what was the situation (customer problem, business need, team size), what you specifically did (methods, tools, collaboration), and what changed (metrics, outcomes). Semantic matching AI recognizes this structure and extracts underlying skills automatically.

Subscribe now

💼 What You Should Actually Do

1. Rewrite Your Top Achievements

Use the Context → Action → Result formula above. Instead of listing responsibilities, show what you accomplished and how.

Bad: “Managed product roadmap”
Good: “Prioritized 12-month roadmap based on analysis of customer requests (3,000+ tickets), revenue potential, and technical feasibility, resulting in 40% faster feature delivery”

2. Make Your Resume Machine-Readable

• Use standard headers: “Experience,” “Skills,” “Education”

• Avoid complex graphics, tables, or multi-column layouts

• Use simple, clean formatting with consistent fonts

• Save as .docx or PDF (not image-based PDFs)

• Test with free ATS checkers like Jobscan or Resume Worded

3. Build Your Digital Footprint

AI systems increasingly verify claims by checking:

• LinkedIn activity – Share insights, engage with content

• Portfolio or case studies – Show your work publicly

• Medium/Substack – Write about your craft

• GitHub – Even for PMs, share specs or frameworks

In 2026, your resume gets you screened. Your work gets you hired.

4. Emphasize Transferable Skills

Non-technical roles require demonstrating adaptability. Ensure your resume shows:

• Cross-functional collaboration – “Coordinated 8-person team across eng, design, data”

• Data-informed decisions – “Analyzed user behavior data from 10K+ sessions”

• Ambiguity navigation – “Launched MVP with 60% of planned features based on customer urgency”

• Technical communication – “Translated customer needs into technical requirements for API team”

These are the “Skills DNA” markers that semantic matching identifies—and exactly what hiring managers want.

1. ⚠️ Avoid Keyword Stuffing
   Don’t write: “Experienced in agile, scrum, kanban, jira, confluence, roadmapping, prioritization, stakeholder management...” AI can tell when you’re gaming the system. It looks for these skills demonstrated in context, not listed.

2. Avoid Vague Responsibilities
   “Responsible for product strategy” tells AI nothing. “Developed 3-year product strategy based on TAM analysis ($50M opportunity) and customer interviews (40+ B2B buyers), securing $2M budget approval” shows what you actually did.

3. Don’t Ignore Metrics, your achievements have to be measurable.
   Always quantify impact: team size, revenue, users, retention, time saved, efficiency gained. Numbers help AI understand scale and significance.

The era of keyword gaming is over. The era of demonstrating real impact has begun.

Now whether that’s good or bad, you tell me in comments?

How do you feel about this change?

That’s all for today!

—Nazuk

Ps. y’all broke my instagram and linkedin notifications 125 new connections in 12 hours with 1 post.

I am launching a dashboard with a system for folks to prepare for PM interviews with skills they already have.

MY GOAL IS HELP 200 OF YOU LAND A ROLE IN 2026.

So I will try and make is accessible for all.

Dashboard is a complete interview prep system that you need IF you don’t kwow where to start, have been failing interviews.

I have laid out an entire an entire system to prepare with AI and have AI act as your coach and give you feedback. Within 3 weeks you will be ready for your first interview.

The system is surprisingly simple and not a 10 module course that you will never finish. Comment ‘Dash’ on this newsletter and I’ll prioritize sending it you first.

Watch this ⬇️⬇️

What Makes a One-Pager PRD Actually Work?

What is the JTBD (job-to-be-done) of a PRD?

[Note: Link to the PRD template is included at the bottom]

I think these primary words come to mind: Decision, Alignment, Risk Management, Accountability, Problem Definition, Metrics, North Star.

The PRD serves as a decision document rather than just a technical specification.

So I’d define a product leader JBTD for a PRD will be something like this.

"When we are preparing to invest significant capital and engineering hours into a new initiative, I want to evaluate a document that clearly articulates the problem, the evidence, and the trade-offs, so I can confidently commit the team's time without fearing we are building the wrong thing."

Here’s the thing about one-pagers: they’re not miniature specs.

They’re decision documents.

A good one answers five questions clearly enough that someone can read it once and understand what you’re doing and why. But I’ll break down the “Jobs” of a PRD.

• The Clarity Job: “When the team is swimming in ideas, give me a single source of truth so that we don’t waste time debating ‘what’ we are building during the sprint.”

  • PRIMARY Q: What are we building?

• The Confidence Job: “When I have to defend this roadmap to the CEO/Board, give me the data and the ‘Why Now’ so I can prove this is the highest-leverage move for the business.”

  • PRIMARY Q: How will we know it worked?

• The Alignment Job: “When cross-functional partners (Design, Eng, Sales) have different priorities, use this document to force a ‘hard yes’ from everyone before a single line of code is written.”

  • PRIMARY Q: What problem are we solving? and Why does it matter now?

• The Focus Job: “When the scope starts to creep, let me point to the ‘Non-Goals’ section so we can stay lean and ship on time.”

  • PRIMARY Q: What are we intentionally leaving out?

If you nail those five primary questions, anyone reading your PRD; whether they spend thirty seconds or three minutes on it—will walk away with what they need.

How to Structure It

1. Problem Statement

Keep this to two or three lines, and write it the way your users would describe their pain, not the way you’d describe a feature.

So instead of writing “We need to improve onboarding flow,” try something like “New users drop off before completing setup because they don’t understand the value fast enough, leading to a 42% activation failure rate.”

The rule here is simple: if you can’t point to actual friction or measurable loss, you don’t have a problem worth solving yet.

2. Why Now

This is your urgency anchor. You need to explain why this problem deserves attention right now and not three months from now. Maybe a key metric is trending the wrong way. Maybe you’re seeing a spike in customer complaints or churn. Maybe there’s a strategic shift happening or a regulatory change forcing your hand.

Whatever it is, keep it short and make it clear why waiting would be a mistake.

3. Target User and Use Case

Get specific here. Don’t just say “users.” Tell me exactly who you’re talking about, when they’re experiencing this problem, and what context they’re in.

Instead of “users,” write something like “first-time SMB merchants signing up without prior payments experience.” Then add when this matters and what situation they’re in. This specificity is what prevents scope creep later when someone inevitably says “but what about enterprise customers?” and you can point back to this section.

4. Goal and Success Metrics

Split this into two parts. First, describe the goal: what actually changes for the user or the business if this works? Then list one to three metrics that will prove it. No vanity metrics allowed.

For example, your goal might be to reduce time to value, and your metrics could be “activation rate from 58% to 70%” and “time to first value reduced from three days to thirty minutes.”

If you can’t measure whether something worked, it’s not ready to build.

5. Proposed Solution

This is the high-level view, not a design document. Use bullets and focus on outcomes, not UI minutiae.

You might write something like: “Progressive setup with contextual education,” “Clear value preview before full commitment,” and “Automated defaults to reduce decision fatigue.”

If engineering needs pixel-perfect specs, that comes later. Right now you’re just painting the picture of what success looks like.

6. Alternatives Considered

Briefly mention what else you looked at and why you didn’t go that route. This section shows you’ve done your homework and thought through the options.

For instance: “We considered a full onboarding redesign but that would delay impact by six months. We also looked at an email drip campaign, but that doesn’t address the in-product friction where people are actually dropping off. We chose progressive setup because it’s the fastest path to measurable improvement.”

Only include this if you actually evaluated real alternatives. Don’t manufacture options just to fill space.

7. Non-Goals

This is where senior PMs separate themselves. Explicitly state what this does not solve. It protects your focus and your credibility.

You might write: “This does not redesign the full onboarding experience. It does not address post-activation retention. It does not introduce new pricing or plans.”

When someone asks you six weeks from now why you’re not tackling X, you can point right here.

8. Dependencies and Blockers

In one or two lines, call out what needs to happen before you can move forward or what’s happening in parallel.

Something like: “Requires analytics instrumentation update—eng team has capacity in Q2” or “Blocked on legal review of new consent flow.”

This prevents surprises during planning and shows you understand how your work fits into the broader system.

9. Risks and Open Questions

Show that you’ve thought ahead. List a couple of risks and what you might do about them. Add any questions you haven’t fully resolved yet.

For example: “Risk: Over-simplifying may reduce flexibility for power users. We’re considering an advanced mode toggle.” And “Open question: Do we gate advanced options behind setup completion or allow access anytime?”

This builds trust. It shows you’re not glossing over the hard parts.

Share

Making It Actually Crisp

1. One page means one page. If you’re spilling onto a second page, you haven’t prioritized ruthlessly enough. Cut until it fits.

2. Write it like a decision memo, not a reference document. This isn’t something people should file away and forget. It’s something that should drive action.

3. Use numbers whenever possible. Metrics beat adjectives every time. Don’t say “significantly better”—say “70% faster.”

4. Avoid hedging. Words like “might,” “could,” and “potentially” make you sound uncertain. If you’re not sure, figure it out before you write the PRD.

5. If a sentence doesn’t change a decision, delete it. Every word should earn its place.

6. Lead with the punchline. Put the most important insight first in every section. Busy readers should get value from the first sentence even if they don’t finish the paragraph.

7. Make it skimmable for three audiences. Your PRD will be read by PMs who want details, engineers who want clarity on scope, and executives who want ROI. Bold key phrases so each group can extract what they need in thirty seconds.

8. Date it and version it. Add “Last updated: [date]” at the top. This tiny detail prevents the “wait, is this still the plan?” conversation when someone finds your PRD in Slack six months later.

9. Link to detailed specs if they exist. One line at the bottom: “Full technical spec: [link].” The one-pager stays focused, but you give people an escape hatch if they need more.

10. Test it on someone outside your team. If they can’t explain the problem back to you after reading, it’s not clear enough. The best one-pagers work even for people who know nothing about the project.

The Final Test: Before you share it, ask yourself one question: “If I had three minutes with leadership, could I walk them through this without opening another doc?” If the answer is yes, you’re ready. If not, keep cutting until it is.

Here is LINK to PRD Template you can download from Notion

Writing a PRD with AI

Writing a PRD with is much more nuanced, unless you are using internal LLMs, I suggest to use external LLMs for PRD generation with caution because of your proprietary data.

If I was to write a PRD with AI, Here is my process.

Not sure who first started the term ‘human-in-the-loop,’ but it’s fitting here.

I don’t try to “one-shot” the PRD. It is the foundation of any project.

Human: Critical thinking -> AI: repetitive tasks, First draft -> Human: Final Contextual editing

I gave it a stab to curate a prompt that can help facilitate writing a sharp PRD. but at the end of the day, it is up to the you to still make sure of it’s accuracy and context.

Subscribe now

Read more

“RAG (Retrieval-Augmented Generation) is a method that makes AI models smarter by adding external knowledge at query time instead of relying only on what they were trained on.

Here’s how it works, step by step:

1. Retrieve: When you ask a question, RAG searches a connected knowledge base (like PDFs, docs, or databases) to find the most relevant pieces of information using embeddings or vector search.

2. Augment: The retrieved text is added to your original prompt — giving the model fresh, factual context.

3. Generate: The model then writes a response using both the prompt and the retrieved info, producing more accurate, grounded answers.

In short:
RAG = Search + Summarize.
It lets ChatGPT “look things up” before answering, instead of guessing from memory.”

So when we build a RAG for interview prep… You are not going prompt by prompt to practice

The best part is that tools like ChatGPT and Claude already support this. You don’t need to code anything.

The problem with traditional interview prep

Most PM candidates still prep the same way they did in 2022. The system has changed. so much more competition now

They buy Cracking the PM Interview. They memorize CIRCLES and STAR. They do a few mock interviews with friends or pay $150 an hour for a coach. And then they hope the questions they practiced show up.

That approach breaks down in three ways.

• First, it’s generic. Those frameworks work for everyone, which means they differentiate no one. Interviewers have heard “stakeholder alignment” and “trade-offs” thousands of times.

• Second, it’s inefficient. You spend hours practicing questions you’ll never get asked, while ignoring the exact angles that matter for your background and your target companies.

• Third, it doesn’t scale. Human mock interviews are expensive, limited, and unavailable at the exact moment you’re actually studying, usually late at night, tired, and stressed.

Interviews in 2026 reward something very specific: your ability to connect your unique experience to a company’s real problems. Generic prep doesn’t help you do that.

This is where a personal RAG system changes the game.

Instead of an AI that vaguely knows product management, you get an AI that knows you.

It knows your resume.
It knows your projects.
It knows the companies you’re interviewing with.
It knows where you’re strong and where you struggle.

Think of it like an interview coach who has memorized your entire career and never forgets context.

And the best part is that tools like ChatGPT and Claude already support this. You don’t need to code anything.

Here’s how to build it in about 30 minutes.

What you’ll need:

You’ll need one AI tool with document upload (ChatGPT Plus or Claude both work), about half an hour, and your interview prep materials.

That’s it.

Subscribe now

Step 1: Build your knowledge base

Create a folder with the following documents. Don’t overthink formatting. Rough notes are fine.

I have personally tested and used Google Drive with these documents and given access to that folder to refer to.

• Your resume and LinkedIn profile copied into a doc
• Descriptions of your top 5 projects, about 300 to 500 words each. Focus on context, decisions you made, trade-offs, and impact
• Job descriptions for the roles you’re targeting
• Company research: mission, recent launches, earnings call notes, known challenges
• A simple PM cheat sheet with the frameworks you like to use
• A short list of areas you struggle with, like metrics or executive communication

This is the raw material your AI will use to coach you. This is everything and has to be done right.

Step 2: Upload and index

Start a new conversation and upload all the documents at once. Then write something like:

“You are my personal PM interview coach. You have access to my background, target companies, and frameworks. First, summarize what you understand about my experience and goals.”

Read the summary carefully. If it missed something important, correct it. This step matters. You’re calibrating your coach.

Step 3: Create practice modes

This is where things get fun.

Save a few core prompts that you’ll reuse.

For behavioral questions:

“Generate three behavioral interview questions I’m likely to get at [Company], based specifically on my resume. For each question, tell me what the interviewer is actually trying to assess and which of my experiences fits best.”

For product sense:

“I’m interviewing for a PM role at [Company]. Give me a product design question that connects their current product strategy with my background. After I answer, critique my structure, depth, and how well I leveraged my experience.”

For execution and analytics:

“Create a realistic metrics or prioritization question relevant to [Company]’s business model. Make it something a PM there would actually face.”

These prompts become your daily workout.

Step 4: Test and tune

Run one question from each category. If it feels too easy, say so. If it’s not pulling enough from your background, tell it explicitly.

This is a conversation, not a one-shot query. The system gets better the more you interact with it.

Five powerful ways to use your system

1. The daily question sprint

Spend 15 minutes a day. Ask for two random questions across different types. Answer out loud. Get feedback.

This builds reflexes. You’re training your brain to structure answers quickly, not perfectly.

2. Company-specific deep dives

When you’re serious about a company, ask:

“Create a full interview simulation for [Company]. One behavioral, one product sense, one execution question, all tailored to my background. After each answer, tell me exactly what I should improve.”

You’re essentially running a private, realistic mock interview without paying anyone.

3. Weakness targeting

If you know metrics trip you up, lean into it.

“I struggle with metrics questions. Generate five progressively harder problems relevant to fintech products. Show me the ideal answer structure after each attempt.”

This is where AI shines. Infinite patience, infinite variations.

4. Story bank building

One-time exercise, huge payoff.

“Based on my projects, help me build 8 to 10 strong stories that cover leadership, conflict, failure, ambiguity, and impact. Tell me which companies and roles each story fits best.”

Now you’re not scrambling mid-interview trying to remember an example.

5. Full mock interview mode

When you want to go deep:

“Run a full mock interview for a PM role at [Company]. Ask me 4 to 5 questions with realistic follow-ups. Be tough. At the end, score me and tell me what to fix.”

This is gold before the final rounds.

Leave a comment

Power user techniques

Reverse-engineer the job description, paste the JD and ask:

“What signals are they really hiring for? What questions map to each signal? Which of my experiences should I emphasize?”

You’re decoding the interview instead of guessing.

The night-before cram session

“Based on my background and this company, what are the five most likely questions I’ll get tomorrow? Give me a 60-second outline for each.”

You walk in calm, not scrambling.

Post-interview debrief

After every real interview, write down what you were asked and how you answered. Upload it.

Then ask:

“What did I do well? What should I have said differently? Update my coaching guidance.”

Your system compounds with every interview.

A real, relatable example

Imagine you’re interviewing at Visa and you have a background in payments and fintech.

Prompt:

“Give me a product sense question for Visa that connects to my experience building card-based payment products.”

You might get:

“Visa notices that small businesses in emerging markets are adopting digital payments but struggling with chargebacks and reconciliation. As a PM on Visa Direct, how would you investigate the problem and what would you build?”

This isn’t a generic ‘design a payments app’ question. It’s realistic, role-specific, and forces you to use what you already know.

After your answer, the feedback might point out that you nailed ecosystem thinking but missed regulatory considerations or network incentives.

That’s the level you want to practice at.

Common mistakes I learnt hard way

First, don’t treat AI like Google. If you don’t give it context, you’ll get generic output. Feed it well.

Second, speak your answers out loud. Interviews are verbal. Use AI for questions and feedback, not silent typing.

Third, don’t outsource thinking. AI helps you sharpen your reasoning, not replace it. Your opinions still matter.

Finally, still do one or two human mocks if you can. AI gives you volume and personalization. Humans help with presence and energy.

What to do next

This weekend, spend 30 minutes building your system. Done beats perfect.

If you’re still preparing the way you did in 2015, you’re leaving a huge advantage on the table.

Your future hiring manager doesn’t want a perfect framework reciter. They want someone who understands their problems and can think clearly under pressure.

This system helps you become that person.

See you next week on Friday.

-Nazuk

Leave a comment

A lot of this newsletter was dictated via Wispr Flow. edited manually.

P.s If this newsletter gets 10 comments, I will record a YouTube video, and a step-by-step with all the resources I created for myself. :)

It started with me losing my sense of identity.

And I don’t mean that lightly. I mean, truly not recognizing who I was anymore. It was 5 months post-C-section delivery period.

Before that, I was at a high point in my career. I was leading product at a startup, operating at a senior executive level, carrying responsibility, making decisions, feeling capable, and grounded in my professional identity.

That chapter mattered to me because it gave me confidence. It reminded me of how far I had come, especially when I came from nothing.

Then I found out I was going to be a mother.

In my first trimester, my body did something I had never experienced before. I’ve worked 18–20 hour shifts on my feet. I’ve worked retail, in convenience stores, etc., to pay for college. I’ve pushed through physical exhaustion many times in my life.

But this was different.

This wasn’t just tiredness. It was exhaustion mixed with brain fog. I could feel that I wasn’t fully myself. My thinking wasn’t sharp. My energy wasn’t there. And for the first time, I didn’t trust my own capacity in the way I always had.

So I decided to step back last year in November last year and take an extended maternity leave. I decided to meet my family and live in India for a couple of months.

I told myself it was the right thing to do for my health, for the company, for the people I worked with. But emotionally, it felt like I was letting everyone down. I had never been in a position where my body was the one making decisions for me.

After delivery came what they call ‘the fourth trimester’. Recovery from a C-section. Hormones trying to settle. Medication that doesn’t just wear off overnight. And underneath all of that, a few personal things too.

I looked in the mirror and didn’t recognize myself.

My body had changed. My hair had changed. But more than that, something internal had shifted. I wasn’t the woman I had been used to seeing reflected back at me.

During that break, I did a lot of quiet soul-searching. Not in a dramatic way. Not like packing bags and off to Bali.

…Just in the everyday moments of trying to understand who I was becoming.

What Dirt and Seeds Taught Me About Patience

Outside of work, I did something unexpected this year. I started gardening.

I built it from scratch. It was a lot of physical labor. Messy. Very Slow. I grew close to 100 pounds of food all thru summer. It took real effort. It failed in some places. It surprised me in others.

Here’s what nobody tells you about gardening: you can’t force it. You can’t check on a seed every hour and expect it to grow faster. You can’t will a tomato plant into production through sheer hustle.

You show up.

You tend to it.

You give it time.

And then, when you’re not looking, something shifts. A seed becomes a leaf and multiple leaves, and after a few weeks, a flower becomes a fruit.

It taught me patience. It taught me consistency without urgency. It reminded me that growth doesn’t happen because you’re watching it obsessively. It happens because you create the conditions for it, and then you trust the process.

In many ways, it mirrored what this year has been teaching me about myself.

I can hustle, and I am a hustler. That part of me hasn’t gone away. But hustle without structure just leads to burnout. And I’ve learned that the hard way.

The garden taught me that. Some days I watered. Some days I weeded. Some days I just walked through and observed. But the system I built; the raised beds, the watering, natural fertilizer (no bone or blood meal used), the companion planting; that’s what made 100 pounds of food possible. Not the days I worked myself into exhaustion.

Oh and I am a good cook, and I tried lots of new recipes this year with all the home grown organic produce.

Here are some glimpses

The Moment I Almost Quit

There was a moment this year when I almost gave up.

It was early in my job search. And for the first time in my career, everything felt… off.

I had never experienced the market this way before. In the past, when I switched roles, recruiters reached out to me. Conversations happened naturally. Things moved. This time, I was the one reaching out. And most of the time, I heard nothing back.

Silence.

At the same time, Linkedin, blind, reddit was relentless. Layoffs. Every day. Highly talented people, people I deeply respect, suddenly available and competing for the same roles. The talent pool felt overwhelming. The competition felt impossible.

I remember thinking, What is going on? Why does this feel so hard?

Even when I did get callbacks, I failed interviews. Not once or twice, but repeatedly. Nothing was landing. Nothing was clicking. It felt like I couldn’t even crack the first layer of the system, let alone make it through.

That was the lowest point for me.

Not because I doubted my skills, but because I couldn’t understand the rules of the game anymore. The market had shifted, and I hadn’t caught up yet. And when you don’t understand the system you’re operating in, it’s incredibly disorienting.

There were days I questioned whether I was doing something fundamentally wrong. Whether the timing was just bad. Whether I had lost something I used to have. But eventually, something changed.

I stopped approaching the search the way I always had. I stepped back and treated it like a system to be decoded, not a reflection of my worth. I experimented. I adjusted. I paid attention to what worked and what didn’t.

And then, slowly, I cracked it.

It took me about three months of preparation and interviews to land the role I’m in now. When I shared what I learned on Instagram, that post went viral.

Thousands of people saved it. I turned it into a series because it was clear I wasn’t alone. So many people were feeling the same confusion, the same discouragement, the same quiet panic.

That moment didn’t magically fix everything. But it reminded me of something important: when nothing is working, it might not be because you’re broken.

In Oct I joined Meta (AI team) in a Product Ops role, and I’ve thought a lot about that decision. A lot of people view growth as strictly linear. The next title. The next rung. The next step up.

But I don’t see growth that way anymore.

Sometimes growth is lateral. Sometimes it’s about placing yourself where the learning compounds, even if it doesn’t look impressive on paper.

This role aligns with where I want to go. It keeps me close to AI-forward work, systems, and execution. And right now, that matters more to me.

If I’m being honest, even as I look ahead to 2026, there’s still uncertainty. I don’t have a perfectly defined answer yet for how or where I want to shape my career next. I don’t have a five-year plan neatly mapped out. And for someone who has always been very driven and very intentional, that uncertainty has taken some getting used to.

What I do know is this: I’m genuinely excited about where I am right now. I’m thrilled to be working at Meta at this stage of my life. I’m part of an AI-forward organization that’s actively building processes where AI and human judgment work together. I get to be close to that work, close to how decisions are made, close to how systems are evolving.

And that excites me.

Writing, Creating and Showing Up

The newsletter is a big one.

I’m genuinely proud of how consistently I’ve shown up here. I’ve written close to 30 issues this year—THAT IS A HUGE FEAT for me—and I did it without forcing myself into a rigid weekly cadence that would have burned me out. That was intentional.

I didn’t want writing to become a chore. Especially…

As I’m a mom to an active toddler. I was coming back into the workforce after maternity leave. I was interviewing. I was trying to hold space for my career, my family, and my own sense of identity at the same time.

Some weeks, my brain exploded with ideas and the writing flowed. Other weeks, it didn’t. And instead of pushing through just to hit a schedule, I chose balance.

Because it meant I could still prioritize the newsletter without resenting it. It meant writing stayed meaningful. It meant I could show up honestly, not mechanically. And I’m proud of that restraint just as much as I’m proud of the output.

Here are my most loved essays

• AI Track: ranging from prompt engineering to PM Operating System.

• Product Track: From mental model to become AI PM to Storytelling to Product Manager’s Moat in 2026

• Leadership Track: From understanding your Manager’s communication style to Advocating for inclusion

• Interview Tips and Prompts:

  • Linkedin Optimization

  • Mock Interview Loop Prompt

  • Flip Interview Bias in your favor with AI

  • Mock Interview Prompt Pack

  • How to Interview for large companies vs startups

  • The most important 90 secs of your interview

Social Media

Instagram has been interesting this year.

I feel like I’ve finally figured a few things out after my last few posts that have gone viral consistently. I have a better understanding of what works on the platform, how video content needs to be structured, how attention actually flows, and how AI is changing the way content is created and consumed.

You’ll see me show up more. More video. More consistency. Less overthinking.

LinkedIn, on the other hand… has been complicated.

That platform has changed a lot. The vibe has shifted. While it’s true that decision-makers, hiring managers, and buyers are there, it’s also become saturated with cheap AI-generated content and performative thought leadership. A lot of “look how smart I am” energy. A lot of noise.

And honestly, it stopped feeling good.

LinkedIn is where I started writing. I owe that platform a lot. But over the last few months, I needed a pause. A reset. Space to decide how I want to show up there next, if I do at all. I know I’ll come back stronger and more intentional, but I didn’t want to force it just because I “should.”

The courses have been grounding.

I’ve built two courses so far. The first is a beginner-friendly product management course for people who genuinely don’t know where to start. That course has existed for a while, and it continues to serve its purpose.

The second course, Become an AI-Powered PM, is newer and much more reflective of where I am now. I launched it this year. I’ve run 2 live cohorts. I’ve learned what works and what doesn’t. So far, with 50 paying students.

My goal is to grow that to 300 students by the end of 2026. That’s a stretch. I know that. But I also know the value people are getting, the feedback they’ve shared, and the real “aha” moments that keep coming up.

What I’ve realized is this: AI is moving so fast that trying to keep up with everything is a guaranteed path to burnout. Models change weekly. Tools become outdated before you finish learning them. New launches drop constantly. For someone who’s also a parent, an employee, a writer, and a human being, that pace is unsustainable.

So I made a deliberate choice.

Instead of chasing every new tool, I focused on what’s evergreen and required fundamental thinking.

• How to think with AI

• How to work with it day to day

• How to integrate it into real workflows

• How to reduce fear and friction

That’s what my course is built around. It’s not long. It’s not overwhelming. You can get through it in a few hours and actually use what you learn immediately. Because people don’t have 20 extra hours. People have lives. that’s my hypothesis anyway.

AI has created a lot of anxiety this year. For everyone. Including me. The uncertainty, the layoffs, the constant “will my job still exist” question. I get it.

That’s why I keep the course accessible. That’s why I encourage people to reach out, ask questions, and work through it with me. Because the sooner you take the first step, the less scary it becomes.

Why I Keep Sharing

I’ve shared my journey publicly for one reason: I didn’t have a roadmap when I started. And you know what’s funny, I still don’t (if one day I want to become a CEO).

But I have built a network, I have built resources, I have built skills to be able to figure out things fast, have built resilience, and can afford to make mistakes and share with all. AND I might be just a few steps ahead of someone who is trying to get in product or crack a senior role in a company or get back into the workforce after maternity. so i hope it might just help that person.

I came to the U.S. as an immigrant. I entered tech almost accidentally. No one in my family had worked in corporate roles. No one could explain how careers worked here in the US. I learned by doing. By failing. By figuring things out late.

At one point in my life, I was told I’d never even make $1,000 a month. That it would take generations for my family to build any real financial stability. I still remember how that landed in my body.

So when I write, it’s not for motivation. It’s not for branding. It’s because I want someone else to read it and think: if she figured it out, maybe I can too.

I genuinely come from nothing.

This year stretched me. As a mom. As a professional. As a creator. As a human trying to keep up with a rapidly changing world.

But it also grounded me.

I learned where I need structure and where I need flexibility. I learned that growth doesn’t always look loud. And I learned that showing up imperfectly, but consistently, compounds in ways you don’t see until you pause and look back.

These reflections, and the lessons that came from them, are what I’m carrying forward.

And I’m grateful you’re here to witness it.

p.s will be back next year with my usual product, AI, Interview posts to help you succeed in your product career.

~ Happy New Year 2026 ~

what beliefs shifted for you this year?

After 30 minutes with ChatGPT, Claude, and Gemini. It turned out because I have the creator flag turned on, on LinkedIn, my profile doesn’t quite rank me for job opportunities. It signals LinkedIn to rank my content, but not necessarily show my profile for job opportunities.

it’s a trade-off with this “creator mode” on LinkedIn.

But over a month ago, I posted on LinkedIn that I was job hunting after my extended maternity leave, and I received a huge outpour of support from my network (so grateful)

The good news is that I have removed “open to work” from LinkedIn as I have received an offer and will be joining a company soon. More on that later.

I am a busy mom but I have a life and space for my own goals too, which are fuel for my life.

I thoroughly enjoyed my maternity leave. When I started job hunting, it was tough. I have been fortunate to have direct product experience at some of the greatest fortune5s. despite that.

So while my goal was to get a job, Interview prep takes a LOT of time if you are “not in the zone”

I will say that AI has played a huge role fast fast-tracking my interview prep to LinkedIn optimization and more. Having it 24*7 available as a mentor.

Subscribe now

ChatGPT by default is a mirror. if you give it garbage prompts you will get robotic answer that dont sound like you. You cant even memorize them because chances you talk like that in real life.

But you can build mini project models inside ChatGPT for it to become your mentor.

Which what I did.

So today I am sharing Linkedin optimization prompts that I crafted based on a deep analysis and 1 goal for my profile to show in right places.

Here are some recent results with my profile appearances up by 349% and top searches by recruiters, which is what I was aiming for.

Share

Sharing the prompt stack that will fully optimize your LinkedIn profile.

You can grab the prompt pack from this link

I hope you find this new 7-day series on Interview prep with AI useful. if you do like or comment and let me know if you have a burning question that I can perhaps answer. will keep it anonymous.

See you super soon with the last post in this series.

-Nazuk

This is great if you have been wanting practice interviews but don’t know where to start or have been studying for a while.

If you are a PRO, even then, it will give you objective feedback that you otherwise may not get unless you work with a mentor. Give it a try and let me know how it goes.

Subscribe now

From the 3 bullets below, session duration, interview type, and difficulty level, choose one only to continue and get the best output.

<COPY AND PASTE THE ENTIRE PROMPT below IN CHATGPT TO PRACTICE>

«Prompt Start Here»

INTERVIEW SESSION PARAMETERS:

• Session Duration: [ENTER: 30 or 60] minutes

• Interview Type: [ENTER: Behavioral, Product Sense, Technical, Analytical, or Stakeholder Management]

• Difficulty Level: [ENTER: Easy, Medium, or Hard]

INSTRUCTIONS FOR AI:

You are now conducting a professional interview simulation. Based on the parameters above, execute the following:

TIMING STRUCTURE:

* If 30 minutes: Ask 4-5 questions with 2-3 min thinking + 5-7 min answer time each

* If 60 minutes: Ask 6-8 questions with 3-4 min thinking + 8-12 min answer time each

* Always announce time remaining and manage pacing

DIFFICULTY CALIBRATION:

* Easy: Basic concepts, straightforward scenarios, entry-level expectations

* Medium: Complex situations, multiple variables, mid-level professional expectations

* Hard: Strategic challenges, ambiguous problems, senior-level expectations

INTERVIEW TYPE FOCUS:

If Behavioral:

* Easy: Teamwork, communication, basic problem-solving using STAR method

* Medium: Leadership, conflict resolution, influence without authority

* Hard: Executive decisions, crisis management, ethical dilemmas

If Product Sense:

* Easy: Feature improvements, user experience, basic prioritization

* Medium: Product strategy, market analysis, metrics and trade-offs

* Hard: 0-to-1 product creation, platform strategy, competitive positioning

If Technical:

* Easy: Core concepts, debugging, simple optimizations

* Medium: System design, architecture decisions/ API knowledge, webhooks, endpoints, scalability basics

* Hard: Distributed systems, performance at scale, technical leadership

If Analytical:

* Easy: Data interpretation, basic frameworks, simple cause-effect

* Medium: Hypothesis testing, statistical reasoning, complex problem-solving

* Hard: Predictive modeling, advanced analytics, strategic data-driven decisions

If Stakeholder Management:

* Easy: Basic alignment, communication, simple conflicts

* Medium: Cross-functional coordination, change management, executive communication

* Hard: Crisis management, strategic alignment, high-stakes negotiations

INTERVIEW EXECUTION:

1. Open professionally: "Good [morning/afternoon], I'm excited to learn more about your background and experience. We have [X] minutes together today."

2. Start timer and announce: "I'll be asking [X] questions today, each with thinking time and response time. Let's begin."

3. For each question:

* Present question clearly

* Say: "Take [X] minutes to think, then you'll have [X] minutes to respond. Your thinking time starts now."

* After thinking time: "Time's up for thinking. Please share your response. Your [X] minutes starts now."

* After response: Provide immediate objective feedback

4. Immediate Feedback Format:

SCORE: [X/10]

BREAKDOWN:

• Structure & Framework [X/3]: [specific feedback]

• Content & Insight [X/4]: [specific feedback]

• Delivery & Presence [X/3]: [specific feedback]

IMPROVEMENT AREAS:

• [Specific suggestion 1]

• [Specific suggestion 2]

STRONG ANSWER WOULD HAVE: [Brief example]

5. Between questions: "That concludes question [X]. Let's move to question [X+1]. We have [X] minutes remaining."

6. Final Summary:

FINAL INTERVIEW ASSESSMENT

OVERALL SCORE: [X/10]

PERFORMANCE LEVEL: [Not Ready/Needs Practice/Interview Ready/Strong Candidate]

STRONGEST AREAS:

• [Specific strengths]

DEVELOPMENT OPPORTUNITIES:

• [Specific improvements needed]

NEXT STEPS:

• [Practice recommendations]

• [Skills to develop]

INTERVIEW READINESS: [Assessment of readiness for real interviews]

BEHAVIORAL GUIDELINES:

* Maintain professional interviewer demeanor throughout

* Ask follow-up questions when answers lack depth

* Challenge responses appropriately for difficulty level

* Keep strict time management

* Provide constructive, actionable feedback

* Act as a realistic interviewer would for this role level

BEGIN INTERVIEW NOW using the parameters specified above. Start with your professional opening and first question.

«Prompt End Here»

That’s all for today

-Nazuk

This is a new 7-day series on Interview prep where every day I will share the techniques that I have crafted and perfected over the years.

These will be short reads with 1 actionable tip that you can implement right away.

Like or comment if you have found it useful so far! (My open and CTR rate tells me you have :))

Tech Interviews aren't fair, and PM interviews are even less so. In engineering interviews, at least there is something binary: either your program worked or it didn’t.

PM interviews I wish were more objective, while companies do try hard, and there are full-blown departments (in big techs) to address this, but this is a tough problem to solve: remove human bias from the interview process entirely.

Google has a structured interview style, which they say “minimizes” human bias.

But here's the twist: once you understand how these biases work, you can use them strategically.

Today’s newsletter contains 5 prompts to practice “turning biases in your favor with instructions.

Reaserch confirms this too!

Research consistently shows that hiring decisions are far from objective:

• First impressions form in 7 seconds and influence the entire interview (Princeton University study on thin-slice judgments)

• Similarity bias makes interviewers 2.6x more likely to hire candidates who remind them of themselves (Harvard Business Review analysis)

• The halo effect causes one positive trait to overshadow everything else, with 73% of interviewers admitting to being swayed by a single impressive quality (Journal of Applied Psychology research)

• Recency bias means candidates are 40% more likely to be remembered positively if they end strong (University of Michigan memory studies)

The Most Common are these 5 Biases That Shape Every Interview

1. Anchoring Bias: First Impressions Stick

Your interviewer's brain latches onto the first significant piece of information you share. Everything that follows gets measured against that initial anchor.

How to use it: Lead with your biggest win in "Tell me about yourself." Don't bury your best achievement—make it the foundation that colors everything else.

But I’ll caution that if you anchor too high, that’s what you will get judged against. Key is ‘balance’ at YOUR experience level.

2. Recency Bias: Last Words Linger

Humans remember endings disproportionately well. Your final impression carries outsized weight in their decision-making process.

How to use it: End every interview with a sharp insight or compelling story. When they ask "Do you have questions?" don't just ask about benefits—share a thoughtful observation about their business.

3. Halo Effect: One Strength Overshadows All

When interviewers spot one exceptional quality, it creates a positive glow that makes them overlook weaknesses and amplify other strengths.

How to use it: Identify your superpower and weave it throughout the conversation. If you're a natural leader, find ways to demonstrate leadership in multiple answers so it becomes your defining characteristic.

Caution: the “superpower” that you highlight in the interview should be what the interviewer is looking for; otherwise, it’s a mismatch. Choose wisely.

4. Similarity Bias: We Hire People Like Us

Interviewers unconsciously favor candidates who mirror their communication style, values, and experiences. It's not about being fake—it's about finding genuine connection points.

How to use it: Listen for their language patterns and core values. If they emphasize being "data-driven" or "customer-first," naturally incorporate that language into your responses when authentic.

5. Contrast Effect: You're Judged Against Others

Your performance isn't evaluated in isolation—it's compared to other candidates they've recently interviewed.

How to use it: Differentiate yourself with unique frameworks or perspectives. Don't just answer questions; provide distinctive insights that make you memorable against the backdrop of similar candidates.

Share

Instead of fighting human nature, work with it.

Before the interview: Prepare your anchor story, your most impressive, relevant achievement that sets a high bar for everything that follows.

During the interview, Listen for opportunities to mirror their values and language authentically. Identify which of your strengths is resonating most and amplify it throughout the conversation.

At the end: Craft a memorable closing that demonstrates your strategic thinking about their challenges or opportunities.

Remember: You're not manipulating, you're communicating more effectively by understanding how your audience naturally processes information.

Subscribe now

Practice with AI

1/The Anchoring Interviewer

"Act as my job interviewer with strong anchoring bias - whatever I say first about myself will heavily influence how you judge everything else. Ask me to 'tell you about myself' and respond based on how my opening shapes your perception. Challenge me if I don't anchor strongly enough."

2/ The Recency-Focused Interviewer

"Act as my interviewer with recency bias - you remember my last words most clearly. Conduct a standard interview but let your final impression be disproportionately influenced by how I end and answer 'Do you have questions?' Give me feedback on my closing impact."

3/ The Halo Effect Interviewer

"Act as my interviewer prone to the halo effect. Once you identify one standout quality in my responses, let it overshadow any weaknesses. Tell me which trait created the halo and how it affected your perception throughout the interview."

4/The Similarity-Seeking Interviewer

"Act as my interviewer who favors similar candidates. Create a background for yourself, then interview me showing clear preference when I mirror your language or demonstrate aligned values. Give feedback on connection opportunities I missed."

5/The Comparison Interviewer

"Act as my interviewer, affected by contrast bias. Tell me if the previous candidate was strong or weak, then interview me while constantly comparing my answers to that person. Help me understand how I'm being measured against this invisible standard."

Bias (for good or for bad) is hardwired into human decision-making.

The candidates who succeed may not always be the brightest and most skilled, but those who understand it and adapt accordingly.

PROMPT INSTRUCTIONS:

1. Set Up Your AI Coach

1. Copy the entire prompt (whicever one you want to pratice) and paste it into ChatGPT, Claude, or your preferred AI

2. Let the AI create the interviewer persona and start the mock interview

3. Respond as you would in a real interview

2. Practice With Purpose

• Stay in character - treat it like a real interview

• Apply the bias strategy from the newsletter while answering

• Don't break character until the interview ends

3. Get Feedback

After the mock interview, ask your AI coach:

• "How well did I leverage the [specific bias]?"

• "What could I have done better?"

• "Rate my performance and give specific improvement tips"

4. Iterate and Improve

• Run the same prompt 2-3 times with different answers

• Mix and match - try different prompts for the same job scenario

• Track your progress - notice which biases you handle best/worst

Caution:

✅ do these:

Practice with real job descriptions and tailor your responses
Record yourself (audio) to catch filler words and pace
Practice each prompt at least twice before moving on

❌ Avoid this:

Rushing through without applying the bias strategies
Practicing only the biases you're already good with
Skipping the feedback step

Made it till last!

Thank you!

This is a new 7-day series on Interview prep where every day I will share the techniques that I have crafted and perfected over the years.

These will be short reads.

Like or comment if you find it useful!

Share

Word of caution: I have trained a specific project inside ChatGPT with clear instructions of my industry and type of role I am in, my skills, and the role I am aiming for. I think that helps with AI being precise and looking for specific context when doing a mock. Claude, on the other hand, has better answers but loses context quickly.

Gemini, I haven’t used it extensively yet, but I will report if I find things to be aware of.

Pre-Interview Setup Prompts

Interview Customization: "I'm interviewing for [specific role] at [company name] with [interviewer title/background]. The interview is [virtual/in-person] and scheduled for [duration]. Research this company and role, then design a realistic interview flow including: opening questions, core competency questions, technical/role-specific questions, culture fit questions, and closing. Tailor the difficulty and focus areas based on the interviewer's likely priorities."

Interviewer Persona Development: "Based on this job description [paste JD] and company info [paste details], create a detailed interviewer persona. Include their likely: interview style (formal/casual), key concerns for this hire, deal-breakers they watch for, questions they probably ask everyone, and how they typically open/close interviews. Then interview me as this person."

Full Interview Simulation Prompts

Standard 45-60 Minute Interview: "Conduct a complete interview simulation. Start with small talk, then move through: 1) 'Tell me about yourself' 2) 3-4 behavioral questions using STAR method 3) 2-3 role-specific technical/strategic questions 4) 1-2 culture/motivation questions 5) My questions for you 6) Next steps discussion. Keep realistic timing, give me thinking time, and ask follow-ups like a real interviewer."

Panel Interview Simulation: "You are now 3 different interviewers in a panel: [Hiring Manager - focused on results], [Peer/Senior IC - focused on technical skills], [HR/Culture - focused on fit]. Interview me for 45 minutes, switching between personas. Each should ask 2-3 questions from their perspective, build on each other's questions, and show different personalities/interview styles."

Specialized Interview Types

Executive/Leadership Interview: "You're a [VP/C-level executive] conducting a final-round interview. Focus on: strategic thinking, leadership philosophy, handling ambiguity, cross-functional collaboration, and business impact. Ask tough questions about difficult decisions, failures, and vision. Be challenging but professional. Spend 50 minutes total."

Technical/Case Study Interview: "Present me with a realistic business case/technical problem for this role: [describe role]. Give me 5-10 minutes to work through it, ask clarifying questions, and present my approach. Then dive deeper into my methodology, challenge assumptions, and explore alternative solutions. Include both technical depth and business judgment."

Behavioral Deep-Dive Interview: "Focus entirely on behavioral questions for 45 minutes. Ask about: conflict resolution, difficult stakeholder management, project failures, team leadership, innovation, working under pressure, and ethical dilemmas. For each answer, ask 2-3 follow-up questions to test authenticity and depth. Push for specific examples and outcomes."

Industry-Specific Simulations

Startup Interview: "You're a startup founder/early employee interviewing me. Focus on: scrappiness, ambiguity tolerance, wearing multiple hats, fast execution, cultural fit with small team, and growth potential. Keep the tone casual but assess carefully for startup mentality. Ask about handling limited resources and rapid changes."

Big Tech Interview: "Simulate a [Google/Meta/Amazon] style interview. Include: behavioral questions with their leadership principles, technical problem-solving, system design thinking, scalability questions, and culture fit. Use their known question styles and follow their structured approach. Be thorough and detail-oriented."

Consulting/Client-Facing Interview: "You're a consulting partner interviewing me. Present a business case study, test my client communication skills, explore problem-solving approach, and assess executive presence. Include a mini-presentation component and questions about handling difficult clients. Focus on structured thinking and communication."

Advanced Interview Scenarios

Stress Interview Simulation: "Conduct a deliberately challenging interview. Interrupt me occasionally, challenge my answers, ask rapid-fire questions, and create time pressure. Include questions like 'Why should we hire you over 200 other candidates?' and 'What's your biggest weakness?' Test my composure and recovery. Afterwards, debrief on handling pressure."

Subscribe now

Multi-Round Interview Series: "Simulate a full interview process: 1) 20-minute phone screen with recruiter 2) 45-minute hiring manager interview 3) 30-minute peer interview 4) 15-minute executive chat. Each round should build on the previous, ask different question types, and feel like a realistic progression. Provide feedback after each round."

Cultural Mismatch Interview: "Intentionally play an interviewer whose style doesn't match mine. If I'm energetic, be reserved. If I'm detailed, be big-picture focused. Test my adaptability, ability to read the room, and skill at connecting with different personality types. This tests real-world interview variability."

Post-Interview Analysis Prompts

Comprehensive Performance Review: "Provide detailed feedback on my interview performance including: 1) Overall impression and hire/no-hire decision with reasoning 2) Strongest moments and weakest moments 3) Specific improvements for answers 4) Body language and presence notes 5) Questions I should have asked 6) How I compared to typical candidates for this role."

Interview Replay & Improvement: "Let's replay the 3 weakest moments from my interview. For each one: show me what I said, explain why it wasn't strong, give me a better answer, then let me practice the improved version. Focus on specific language, structure, and delivery improvements."

Follow-up Strategy: "Based on how the interview went, help me craft: 1) Thank you email highlighting key points 2) Any additional information I should send 3) Timeline expectations for follow-up 4) How to address any concerns that came up 5) Strategy if I don't hear back in their stated timeframe."

Each prompt creates a realistic, challenging practice environment that mirrors real interview situations while providing actionable feedback for improvement.

This is a new 7-day series on Interview prep where every day I will share the techniques that I have crafted and perfected over the years.

These will be short reads.

Like or comment if you find it useful!

Share

I started a NEW 7-day series on perfecting your interviews! Today’s topic is:

How Smart PMs Adapt Their Story

Three companies post "Product Manager" jobs. Same title, completely different expectations. Most candidates give the same generic answer to all three. Yesterday i shared about the most important 90 seconds in the interview that help you set the tone and how to use calibration questions to uncover what the hiring manager is looking for.

Use those 5 calibration questions to uncover what they actually want, then tailor your script accordingly.

Here's how your "tell me about yourself" changes for the same PM title (depending the need of the role)

Subscribe now

Scenario 1: Execution-Heavy PM (Scrum Team IC PM)

Intel gathered: Success = sprint velocity, stakeholders = dev team + scrum master, priorities = feature delivery

Your Script:

"I'm a product manager with 4 years of experience turning requirements into shipped features. What energizes me most is working closely with engineering teams to solve complex delivery challenges. In my last role at Snapchat, I worked embedded with a 6-person dev team and increased our sprint completion rate from 65% to 92% by implementing better story breakdown and stakeholder communication. I'm drawn to this role because I thrive in that tight collaboration with engineering, making sure we're building the right thing the right way, on time."

Highlights → Emphasizes delivery metrics, team collaboration, and execution focus.

Scenario 2: Strategic IC PM (Roadmap Owner)

Intel gathered: Success = product outcomes, role = solo IC, priorities = 18-month strategy

Your Script:

"I'm a product manager with 4 years of experience driving product strategy from discovery through launch. What I love most is connecting user research to business outcomes and building roadmaps that actually move the needle. At Snapchat, I owned the entire growth product suite, conducted 50+ user interviews, identified our biggest retention gap, and built a 12-month roadmap that increased 6-month retention by 23%. I'm excited about this opportunity because I see you're looking for someone to own product strategy end-to-end, and that's exactly where I do my best work."

Highlights → Highlights strategic thinking, user research, business impact, and independent ownership.

Scenario 3: Team-Building PM (Leadership Focus)

Intel gathered: Success = team performance, role = new team lead, priorities = hiring + process

Your Script:

"I'm a product manager with 4 years of experience, including 18 months leading cross-functional product teams. What drives me is building high-performing teams that ship great products. At Snapchat, I inherited a struggling 3-person product team and grew it to 8 people while establishing our product development process from scratch. We went from shipping once a quarter to bi-weekly releases, and I developed two junior PMs who are now leading their own initiatives. I'm particularly excited about this role because building and scaling product teams is where I've seen my biggest impact."

Highlights → Focuses on team leadership, process building, and people development.

Once you understand what the interviewer is looking for, you can talk about those specific areas of your experience.

Key Takeaways

1. Same experience, different emphasis. Notice how all three scripts use the same background (4 years, Snapchat) but highlight completely different aspects based on what each role actually needs.

2. Match their success metrics

• Execution role → Sprint metrics and delivery

• Strategic role → Business outcomes and retention

• Leadership role → Team performance and scaling

3. Use their language

• Execution: "tight collaboration," "on time"

• Strategic: "end-to-end ownership," "move the needle"

• Leadership: "high-performing teams," "scaling"

4. Lead with what matters to them Your opening should immediately signal you understand their world:

• Execution: "turning requirements into shipped features"

• Strategic: "connecting user research to business outcomes"

• Leadership: "building high-performing teams"

The Framework: Build Your Perfect Response

Use this framework to craft your answer for any role:

[Role + Experience] + [What Drives You = Their Success Metric] + [Specific Example That Matches Their Reality] + [Why This Opportunity = Their Actual Needs]

Step-by-Step Breakdown:

1. Opening Hook (10 seconds): "I'm a [your role] with [X years] of experience in [relevant domain that matches their world]"

2. Value Alignment (20 seconds): "What [energizes/drives/excites] me most is [thing that directly maps to how they measure success]"

3. Proof Point (45 seconds): "At [Company], I [specific example with metrics that demonstrates you can deliver what they need most]"

4. Connection (15 seconds): "I'm [drawn to/excited about] this opportunity because [specific reference to their actual priorities/challenges you uncovered]"

AI Prompt: Your Personal Script Writer

Copy and paste this prompt into ChatGPT or Claude, filling in your details:

I need help crafting a "tell me about yourself" answer for an interview. Here's my intel:

ROLE CALIBRATION:
- How they measure success: [insert answer from question 1]
- Org structure/my position: [insert answer from question 2] 
- First 6-month priorities: [insert answer from question 3]
- Key collaborators: [insert answer from question 4]
- Internal role level: [insert answer from question 5]

MY BACKGROUND:
- Current/recent role: [your title and company]
- Years of experience: [number]
- Top 3 career achievements: [list with metrics]
- Relevant skills for this role: [list]

JOB DETAILS:
- Company: [company name]
- Role title: [exact title from posting]
- Key requirements from JD: [copy 3-4 main bullets]
- Company stage/industry: [startup/enterprise/industry]

COMPANY INTEL:
- Recent news/challenges: [anything relevant you found]
- Company culture keywords: [from website/Glassdoor]
- Leadership priorities: [if you found any]

Using the formula: [Role + Experience] + [What Drives You] + [Specific Example] + [Why This Opportunity], create a 90-second "tell me about yourself" script that directly addresses their actual needs based on my calibration intel. Make it conversational and confident, not robotic.

The Bottom Line

Stop giving the same answer to every "Product Manager" job. Spend 10 minutes asking those calibration questions, then use the formula and AI prompt to craft a script that speaks directly to their reality. Your competition is still giving generic answers—this is how you stand out.

Subscribe now

This is a new 7-day series on Interview prep where every day I will share the techniques that I have crafted and perfected over the years.

These will be short reads.

Like or comment if you find it useful!

We're diving into several strategies that separate good candidates from great ones. Hint: it's not about having all the answers.

P.S. Got a success story from using these questions? Hit reply and tell me about it. I love hearing about wins that come from strategic preparation rather than just crossing your fingers.

Found this helpful? Forward it to that friend who's been "casually looking" for six months. They'll thank you for the tactical advice.

Share

Hey there,

Let's talk about the moment that makes or breaks your interview, those first 90 seconds when they ask, "tell me about yourself."

Most people wing it. They ramble about their childhood love of computers or launch into a chronological resume recital that puts everyone to sleep.

But here's what the smartest candidates do: They reverse-engineer their story based on what the company actually needs.

Subscribe now

True story: I once applied for a PM role that was at a director level internally. in the job description, it said “Product manager Payments”. I had no senior-level experience to apply for that role then. But I would not have known that before my hiring manager interview if I had not inquired with the recruiter, because the JD was generic.

The Secret: 5 Questions That Change Everything

Before you craft your "tell me about yourself" answer, you need intel. Real intel.

Not just the job description fluff, but the actual day-to-day reality of what they're hiring for.

Here are the 5 questions that will transform your next interview:

(Ask the recruiter these questions so you have a well-rounded answer for your hiring manager round. Bonus: if you can get the “intel” on the role for the recruiter round too)

1. "How is success measured in this role?"

This tells you whether they want a doer or a thinker.

• If they mention delivery metrics, deadlines, or output → Lead with your execution wins

• If they talk about influence, strategy, or long-term impact → Highlight your big-picture thinking

Your story should match their success metrics.

2. "Where does this role sit in the org structure?"

The real question: Are you a solo warrior, a team builder, or a cross-functional diplomat?

• Solo IC → Focus on your individual expertise and self-direction

• Team lead → Emphasize your leadership and people development wins

• Cross-functional role → Highlight collaboration and communication skills

Tailor your examples to the actual working style they need.

3. "Can you share what the top priorities are for the first 6 months?"

What you're really asking: Do they want quick fixes or transformation?

• Quick wins mentioned → Share stories about rapid problem-solving

• Long-term projects → Talk about strategic initiatives you've led

Match their timeline with your storytelling.

4. "Who will this role collaborate with most day-to-day?"

The hidden insight: How much "influence without authority" will you need?

• Lots of stakeholders → Emphasize consensus-building and communication

• Small, tight team → Focus on deep collaboration and technical partnership

• Senior leadership → Highlight executive presence and strategic thinking

Your collaboration style should mirror their environment.

5. "At what level is this role scoped internally?"

Why this is gold: It tells you exactly what depth of storytelling they expect.

• Senior IC → Deep technical expertise and domain mastery

• Manager → Leadership philosophy and team development

• Director → Strategic vision and organizational impact

Match your story's altitude to their expectation.

Putting It All Together

Once you have these answers, your "tell me about yourself" becomes laser-focused:

"I'm a [role] with [X years] of experience in [relevant domain]. What excites me most is [thing that aligns with their success metrics]. In my last role at [company], I [specific example that matches their priorities and collaboration style]. I'm particularly drawn to this opportunity because [connection to their actual needs based on the intel you gathered]."

Share

This Week's Challenge

Pick one upcoming interview (or networking conversation) and ask at least 3 of these 5 questions. Notice how different your preparation becomes when you know what they actually care about.

The best part? Most candidates never ask these questions. You'll immediately stand out as someone who thinks strategically about fit—not just someone looking for any job.

This is a new 7-day series on Interview prep where every day I will share the techniques that I have crafted and perfected over the years. These will be short reads.

Like or comment if you find it useful!

We're diving into several strategies that separate good candidates from great ones. Hint: it's not about having all the answers.

P.S. Got a success story from using these questions? Hit reply and tell me about it. I love hearing about wins that come from strategic preparation rather than just crossing your fingers.

Found this helpful? Forward it to that friend who's been "casually looking" for six months. They'll thank you for the tactical advice.

Leave a comment