In one sentence
Product/market fit is not discovered by building in isolation or following a guaranteed formula. It is more likely when founders combine a personal competitive edge with a plausible market theory, test that theory with customers, ship quickly, sell aggressively, improve retention, and repeatedly confront uncomfortable evidence.
Overview
Cohen frames PMF as a company that is growing sustainably because customers want to pay and continue paying. His roadmap begins before product development: understand why you are suited to solve a particular problem, assess whether the market is large, urgent, funded, reachable, and durable, and identify a sharply defined ideal customer profile. Then build and ship an SLC—simple, lovable, and complete—so real usage can replace speculation. Growth requires deliberate marketing and sales, not merely more features. Product work should primarily improve retention, while a strict milestone-and-obstacle system protects scarce founder time. The final requirement is psychological: tolerate rejection, uncertainty, changing direction, and evidence that contradicts what you want to believe. Cohen contrasts WP Engine, which followed the roadmap closely, with Smart Bear, which found its successful product through a less orderly sequence of building, selling, observing misuse, and pivoting. The roadmap is therefore a repeatable discipline, not a guarantee or universal sequence.
Core ideas
1. Personal fit is more than passion
Passion supplies persistence, but it does not create a business or a defensible advantage. The founder needs distinctive leverage: an unusual combination of skills, experiences, taste, and interests that makes them unusually well suited to a particular path. The key question is: why are you the right person to build this company?
2. Start with a market theory
A product idea is only a hypothesis about customers, competition, urgency, budgets, purchasing behavior, and durability. Cohen recommends testing whether the market is plausible in size, self-aware of the problem, lucrative, liquid enough to buy now, eager to buy from you, and enduring enough to keep paying. Write down the theory so reality can disprove it and guide a deliberate pivot.
3. Customers are the decisive source of evidence
Talk with potential customers before investing months in construction. Interviews reveal how people describe the problem, what they do today, how budgets and purchasing work, where they look for solutions, and what triggers a buying decision. Positive statements about a hypothetical product are weak evidence; clear rejection can save enormous time. Use interviews to discover a narrowly defined ICP whose needs and language can shape the product, positioning, and pricing.
4. Ship an SLC, not a broken MVP
The first product should be Simple, Lovable, and Complete. Simple enables fast learning; lovable gives customers a reason to tolerate limitations; complete means it actually performs a coherent job rather than exposing an unfinished fragment. The important transition is from founder speculation to observing customers attempt to use—and fail to use—the product.
5. Treat marketing and sales as core product work
Founders naturally retreat into coding and design because those activities feel controllable. Cohen’s daily growth questions are: how can more target customers reach the site, how can more try, and how can more buy? The answer may involve product changes, but often lies in messaging, distribution, sales conversations, partnerships, pricing, or conversion work outside the product.
6. Retention tests whether the promise is real
Acquisition and payment show that the promise was attractive; retention shows whether the product delivered it. Low retention cannot be permanently repaired by adding more top-of-funnel growth: growth eventually slows, and customers who leave are evidence of poor fit or poor execution. Investigate unsuccessful users while they are still struggling, compare them with successful users, and improve onboarding, workflows, integrations, product behavior, or ICP selection.
7. Prioritize around one milestone and one obstacle
Choose the next concrete milestone, identify the uncomfortable obstacle preventing it, and select one major initiative—the Rock—that attacks both. Handle only a few high-value Pebbles and let most Sand be merely good enough. This prevents attractive but irrelevant work from consuming the time needed to solve the company’s existential problem.
8. The psychological work is part of the method
The process demands facing facts that threaten identity: customers may reject the idea, competitors may be better, the market may be wrong, and the founder may need skills or people they do not possess. Success requires making specific choices while uncertain, distinguishing useful rejection from noise, and changing direction without dissolving into random activity.
Practical takeaways
- Write a one-page theory of how the company will win: target customer, painful problem, current alternative, willingness to pay, distribution, differentiation, and why the market is large and durable.
- Define an ICP narrowly enough that the homepage, language, pricing, and sales pitch feel unmistakably designed for that customer.
- Before building substantially, conduct structured conversations focused on recent behavior, existing alternatives, budgets, buying triggers, and consequences of not solving the problem.
- Ship a small but coherent product that delivers one meaningful job. Do not confuse a short feature list with a usable product.
- Set a daily question: what action today will increase qualified traffic, trial, conversion, or retention?
- Track cancellation and signs of customer struggle. Contact users before they disappear, and compare unsuccessful users with successful ones to find actionable differences.
- For every period of work, name one milestone, one true obstacle, and one Rock. Treat everything else as secondary unless it directly supports them.
- When progress stalls, ask whether you are doing work because the business needs it or because you are comfortable and skilled at it.
Caveats and counterpoints
- Cohen explicitly rejects the roadmap as a fail-safe formula. Smart Bear reached its successful product through an initially weak personal project, customer-driven discovery, and multiple pivots; a rigid interview-first process might have stopped that path before the useful insight emerged.
- Customer interviews are informative but imperfect. People may not know what to request, may overstate hypothetical willingness to pay, or may lack the vocabulary to describe an opportunity. Actual product usage remains crucial.
- Retention is a lagging, multi-factor signal. It is valuable as a high-level PMF check but is not sufficient as a day-to-day operating metric; leading indicators of customer success are needed.
- The article’s examples are heavily shaped by software startups, especially SaaS. Its advice may transfer less directly to businesses with long procurement cycles, regulatory constraints, network effects, hardware dependencies, or products purchased infrequently.
- The claims about market size, growth, retention, and the WP Engine trajectory are Cohen’s account and examples, not a controlled test proving that every step causes success.
Questions worth revisiting
- What specific combination of experience, skill, access, or taste gives me an edge in this market?
- Which assumptions in my market theory would be fatal if false?
- Who is the narrowest customer segment for whom this problem is urgent, funded, and already being addressed somehow?
- What evidence would distinguish polite interview enthusiasm from real willingness to buy or switch?
- What is the smallest complete product that can expose real usage and failure quickly?
- What promise does my marketing make, and what observable behavior would show that the product is delivering it?
- What is the current milestone, and what uncomfortable obstacle is genuinely blocking it?
- Am I persevering on the right work—or merely becoming more efficient at avoiding the work the business needs?
Return to this when…
Return to these notes when you are tempted to build another feature without customer evidence, when acquisition looks promising but retention is weak, when your roadmap has become a collection of unrelated tasks, or when “perseverance” may actually mean continuing in a direction that evidence has already weakened. The central reminder is to build, sell, observe, and change based on reality—not preference.