The before-and-after moment when a market starts pulling your product out of your hands instead of you pushing it on them.
The phrase "product-market fit" was coined by Marc Andreessen in his 2007 essay The only thing that matters, part of the Pmarca Guide to Startups. His original definition is the one worth memorizing:
Research-based overview. This article synthesizes Andreessen's original essay, Sean Ellis's 40% survey methodology, and Rahul Vohra's product-market-fit engine framework. How we research.
Andreessen's definition is intentionally vibes-based: "you can always feel it." That's frustrating if you're a founder asking "do I have it?" But it's honest. PMF is a discontinuity in the experience of running your business, not a number on a dashboard. The dashboard catches up later.
Three frameworks have stuck. Founders use them interchangeably without realizing they're different.
The qualitative original. The signs are: customers buy as fast as you can build, support tickets are about wanting more not asking why, your inbox fills up unbidden, and you're hiring help just to keep up. The mirror image — pre-PMF — is also recognizable: deals stall, customers ghost, you're always pushing, and growth requires constant founder hustle.
Sean Ellis (early growth at Dropbox, LogMeIn, Eventbrite) operationalized PMF into a single survey question:
If 40% or more of your active users answer "very disappointed," you have PMF. The threshold came from observing companies that grew sustainably and those that didn't; 40% was the cutoff Ellis found correlated with later success. Below 25%, you almost never have it. Between 25 and 40%, it's ambiguous — usually a signal that some users love you but the market is too narrow.
The Ellis test is useful because it's falsifiable. You run the survey, you get a number. The catch: the survey only works on engaged users, and you need at least 40–50 respondents for the result to mean anything.
Vohra (founder of Superhuman) extended Ellis's test into a continuous improvement loop. After running the survey, segment respondents who answered "very disappointed" and ask them: what is the main benefit you receive from the product? Then separately segment respondents who answered "somewhat disappointed" and ask: what would make you very disappointed if it were missing?
The first segment's answers tell you what to double down on. The second segment's answers tell you what feature gap is keeping them off the love-it list. You build that, re-survey, and watch the "very disappointed" percentage climb. Vohra's team at Superhuman moved theirs from 22% to 58% over about 9 months using this engine.
Founders chase abstract definitions and miss concrete observations. Here are three signals that almost always correlate with PMF, and that you can check without surveys.
Notice none of these are revenue numbers. PMF and revenue correlate but they're not the same. You can have $50K MRR without PMF (a high-touch sales motion grinding out deals from a market that doesn't actually want the product) and you can have PMF at $2K MRR (a small, intense user base who love you and would tell their friends if there were more of them).
Founders pattern-match early money to PMF. Here are the most common ways this goes wrong:
The hardest part of PMF is acknowledging when you don't have it. Founders rationalize because the alternative is admitting they need to change something. Andreessen's test — "can I feel it?" — works because if you have to ask, you probably don't. Our piece on validating before building goes into how to test demand before you waste 6 months on a product the market doesn't want.
The pre-PMF rules are short and brutal:
The post-PMF rules are different:
The honest answer: longer than you think, and almost always months not weeks.
Public data on bootstrap solo SaaS — from Indie Hackers interviews, MicroConf talks, and Twitter retros — suggests the median time from "started building" to "clear PMF signals" is somewhere between 12 and 36 months. Outliers exist (some founders find PMF in 6 weeks, especially when riding a hot wave like AI) but the typical path involves at least one or two pivots, multiple landing-page rewrites, and several niche refinements.
This is not because solo founders are bad. It's because PMF is a search problem, and search takes iterations. The pattern in our case studies of micro-SaaS founders: most of them spent 12+ months on something that didn't work before stumbling onto something that did. The ones who found PMF fast usually had unfair advantages (deep domain expertise, an existing audience, a unique technical insight). Most don't. That's normal. The rate of progress is the wrong question; the right question is "am I closer to PMF now than I was 90 days ago?"
If you're in the search phase, keep your burn low, stay alive, and don't confuse activity with progress. The path from zero to $1K MRR is mostly about narrowing your wedge until the market starts pulling.
Product-market fit is the single most important concept in startups, and the hardest to operationalize. Andreessen's "you can feel it" test is the most accurate; Sean Ellis's 40% survey is the most measurable; Rahul Vohra's engine is the most actionable. Use all three. The most important rule is the negative one: do not scale, hire, or raise before PMF, because each of those decisions assumes a fit you don't yet have. After PMF, the rules invert and you should do all three. The frustrating truth for solo founders is that finding PMF takes 12–36 months on average, requires staying alive long enough to iterate, and rewards patience more than effort. The founders who succeed are the ones who can afford to keep searching.
The stack, prompts, pricing, and mistakes to avoid — for solo founders building with AI.