"I'd Know Why We're Building What We're Building"
A PM I'd taught the year before came to coffee. She'd been a senior PM at a fast-growing fintech for two years, running a product area with five engineers and a designer. She was leaving — not to another job, but to build her own product nights and weekends with Cursor and a domain she'd registered three weeks earlier.
I asked the obvious question. "What would you build?"
She paused. "I'm not sure yet. I have some ideas."
I asked the second question. "Then why are you leaving?"
She said something I'd heard from maybe forty PMs in the previous twelve months, almost word for word: "Because at least in my own product, I'd know why we're building what we're building."
I'd been running a regular survey across the people I teach — several thousand product builders go through these courses every year, across all stages and most disciplines. The numbers kept saying the same thing. More than half of them were in frustration, burnout, or open anger about their work — and it wasn't the workload, they were used to long hours. It was something subtler: the sense that the work didn't add up to anything.
Different feature this sprint, different focus next quarter, a new segment the CEO mentioned on Monday that nobody had heard about on Friday, a customer complaint that became a roadmap item without anyone deciding it should. They worked hard. The work didn't add up.
The dream they all shared — "I'll build my own product" — was less about the product and more about the place behind it, where focus would finally exist.
It took me thirteen years to see what sits underneath this dream: a mistake every product company is making in plain sight.
The mistake starts with a belief most teams never say out loud: strategy and execution are different kinds of work, done at different times. Strategy is what happens at the off-site, in the deck, with the leadership team in the room. Execution is everything else: tickets, landing pages, pricing tiers, sales scripts, new hires, the metric on Monday's dashboard. "We're not doing strategy this quarter — we're just shipping. Strategy is what we'll figure out when we have time."
The previous chapter showed how a wrong choice of Job and segment breaks everything below it — value, communication, unit economics, acquisition, conversion. This chapter is about how that choice gets made. It doesn't wait for the off-site. It happens a thousand times a sprint, by nobody in particular, while the team feels like it's just shipping.
Every Decision Is a Strategy Vote
Every product decision the team makes has already chosen which Job of which segment it serves. A ticket, a price tier, a landing page — each of them serves somebody specific. If nobody asks "who is this for?", the answer gets picked anyway: by the loudest customer in the last interview, by the cheapest channel in the ad account, by the customer who emailed the CEO on Monday.
Hiring is the vote almost nobody marks. Say the team hires a senior account executive who spent ten years selling to enterprise. Six months later the pipeline is full of enterprise deals — she sells to the customers she knows, the way she knows how. Nobody decided to go upmarket. The hire decided it.
Each of these votes looks like common sense. "We should answer the email from the CEO's friend." "Sales is closing a deal — they need SSO." "Conversion dropped — let's go talk to people who didn't sign up." The strategy hides behind common sense, every single time. That's why nobody sees it for years. I didn't see it for thirteen.
A simple check for Monday morning: take the last ten decisions the team made — tickets, features, pricing changes, hires. For each one, try to name which Core Job, of which segment, and which success criterion it improves. Count how many answers come back "unclear" or "for everyone." That count is how much of your strategy is happening by accident.
The Order of Your Filters Quietly Decides the Year
A team sets up an interview round. The brief says: "Five conversations with users who started the signup flow but didn't complete it. We want to understand what's missing in our onboarding." Nobody at the table thinks they're making a strategic decision. They're filling a research backlog.
The brief's first filter is conversion. It just cut the market into three slices: people who signed up and stayed, people who signed up and left, people who never signed up at all. The interviews will hear from the middle slice. Every later filter — what Jobs people perform, what they pay, industry, role — applies only inside that middle slice. The other two slices are gone from the analysis. They won't come back, because there is no slot for them in the recruiter brief.
The interviews will explain a lot about why the middle slice didn't convert. They will say nothing about the people who never signed up — and for most products, that is the largest and most attractive part of the market.
I've run this exercise in workshops dozens of times. Every time, the same thing happens in the room. Somebody volunteers a recent research project. They describe it in a sentence. Within two more sentences, they realize that the first filter they applied — "talk to churners," "talk to paying users," "talk to enterprise prospects" — quietly removed the slice of the market where the best segment for their product was most likely to live. The choice of first filter was a strategic decision. They never marked it as one.
The danger lives in the order of the filters, not in any one filter. Reorder the filters and the target segment changes. Apply revenue first, then Jobs — you get one segment. Apply Jobs first, then revenue — you usually get a different segment, often a better one. The same criteria, reshuffled, produce a different strategy.
The worst part is that the order is rarely a real-time decision. It's usually written into a template — the default recruiter brief, the default funnel definition, the default CRM view, the default question order in the user survey. Templates are fossilized strategic decisions, made by nobody, applied by everybody. Every team has them. Almost no team has audited them.
A second check, also for Monday: open the last recruiter brief or analytics filter the team ran. Count how many filters come before the first mention of a Core Job. If the answer is more than zero, the market is being cut before anyone has seen it. Then flip the order in your head and ask who would show up now that didn't before. Those people are the segment the brief silently removed.
Forty Decisions, Voting in Forty Directions
A hidden choice is bad enough on its own. What makes it unbearable: the hidden votes usually point in different directions.
A feature this sprint serves the enterprise security buyer the sales team is chasing. Next sprint's landing page speaks to the indie hacker who organically signed up. A pricing change two weeks later calms the mid-market account that gave the CEO a beating about overage costs. The ad account, meanwhile, is running creatives tuned to whoever clicks cheapest. A support template the week after serves the legacy plan the team is officially deprecating. Every individual decision is reasonable. Together they vote for five different strategies at once, inside the same product team. The team that ships them feels, by the end of the quarter, that the quarter didn't add up.
This is what the people leaving for vibecoded pet-projects are running from. The dream of "my own product" is the dream of a place where the same hand makes every decision, and every decision points the same way. The team is tired from work that doesn't add up.
People feel it long before the dashboard shows it. Engineers and PMs notice the mess before the metrics do. Pull a trusted senior engineer into a 1-on-1 and ask: "how would you describe our product strategy in one sentence?" If the answer is a long pause, a shoulder shrug, or "well, we're doing X, and also Y, and there's Z" — the mess is already there, eating morale months before the retention curves catch up.
Across the people I've watched build products for the past decade, the ones who stayed at one company for a long time had one thing in common. The team had a clear answer to "why this Job, why this segment, why us." The answer survived last week's customer email, last quarter's board mention, and last month's competitor launch. The ones who left, almost without exception, left a place where the answer didn't exist or kept changing.
That's what the survey numbers were saying: over half of several thousand product builders a year. The dream of "my own product" is the visible part. Underneath is a cost nobody ever counted.
Going solo doesn't fix this. A solo builder casts the same votes: a landing page copied from the competitor who seems to be winning, a pricing tier lifted from a template, a feature shipped because the loudest thread on Reddit asked for it. The same hand makes every decision — and still doesn't know what it's choosing. The PM from that coffee will meet the same trap inside her own product within a year. What fixes it is naming the choice — solo or not.
Wes Says We'd Ship Nothing
Strategy is what the team actually does, not what it declares. If what the team does — ticket by ticket, hire by hire — serves the Core Jobs of one segment, most of the win is already in. The line turns a strategy chosen by accident into one chosen on purpose. It costs nothing. Not adding it costs a quarter, then two, then the year — and the senior engineer who walks out because she can't tell anyone what the strategy is.
The Other Side of the Table
Every vote so far chose a customer's Job — but the business on the other side of the table has Jobs of its own, and that's where the next chapter goes.