Every conversation with a user, every abandoned cart, every feature nobody touches is data. The founders who reach product-market fit fastest are rarely the ones with the cleverest idea. They are the ones who built the tightest loop between shipping something, watching what happens, and changing what they build next. This lesson shows you how to construct that loop, how to separate real signal from politeness, and how to turn each cycle into a concrete next step.
Build the loop tight and short
A feedback loop has four stages: ship a small change, put it in front of real users, observe and ask, then decide the next change. The single biggest lever is speed. A loop that turns weekly teaches you far more in a quarter than one that turns once. So ship the smallest version that produces a real reaction, not a polished product. Talk to five to ten users per cycle, not fifty. You will hear the same two or three problems repeat, and that repetition is your signal.
Watch what they do, not just what they say
People are wired to be encouraging, especially in India, where a first user is often a friend, a cousin's contact, or a warm introduction who does not want to disappoint you. Ask "Would you use this?" and almost everyone says yes, and almost none of it is true. Rob Fitzpatrick's book The Mom Test gives a simple fix: ask about their life and past behaviour, never about your idea. Instead of "Would you pay for this?" ask "What did you do the last time you faced this problem, and what did it cost you?"
Then weigh what people actually do above what they claim. A user who says the price is fine but never completes the UPI payment has told you the truth with their behaviour. Track your real signals: where people drop off, whether they return unprompted, and whether they use the product a second and third time.
Measure the few signals that matter
You do not need a dashboard of fifty metrics. Focus on three.
- Cohort retention. Group users by the week they joined and track how many come back over time. Andrew Chen lists a retention curve that flattens, rather than bleeding to zero, as a core sign a product has fit. If every cohort drains away, no amount of marketing will fix it.
- The Sean Ellis test. Introduced by growth expert Sean Ellis in 2009, it asks users a single question: "How would you feel if you could no longer use this product?" If 40 percent or more answer "very disappointed," you are close to product-market fit. Below that, keep iterating.
- Net Promoter Score. Fred Reichheld's 2003 Harvard Business Review article, "The One Number You Need to Grow," proposed asking how likely someone is to recommend you. Treat the number as directional, and pay closest attention to the written reasons behind it.
Keep collection cheap. WhatsApp is the most widely used messaging channel in India and an honest, low-friction way to ask three quick questions or set up a short call.
Turn feedback into the next iteration
Feedback that does not change what you build is just entertainment. After each cycle, sort what you heard into three buckets: patterns you heard from many people, one-off requests worth parking, and noise to ignore. Then pick a single problem, the one that is both most frequent and most painful, and make it the only goal of the next build cycle. Write down the change, the metric you expect it to move, and the date you will check. Close the loop by telling the users who raised the issue that you acted on it. That turns respondents into a standing panel who keep teaching you, cycle after cycle.
Your loop in one line
- Ship the smallest change that creates a real reaction.
- Ask about past behaviour, not future intentions.
- Trust the drop-off data over the compliment.
- Each cycle, change exactly one thing and name the metric it should move.