The market is not the buyer

At Subziwalla, I could describe the market in one breath: Indian and South Asian grocery delivery. It sounded specific. It still did not tell me what to build next.

Take two people who fit that label. One is in the kitchen trying to finish a first order before dinner. The other shops every week, gets a substitution, and does not want to start over. Those are illustrations, but the product choice between them is real: making a first order easier and making a recurring order less annoying are different jobs. A team can spend a month improving one while the customer they actually need is struggling with the other. The market label will sit there looking helpful the entire time.

To be fair, the label does help. It tells you where to look, which constraints may matter, and what you need to understand before pretending you understand the customer. That is a useful start. It just cannot finish the decision.

Now I try to make the conversation more specific, which is also when it becomes more annoying. Who is the person? What are they trying to do at that moment? What workaround are they already tolerating? And what gets worse if they leave the problem alone? That last part matters because two features can both sound reasonable until one of them is attached to a missed sale, wasted hour, embarrassing handoff, or another person who has to step in. Then the roadmap starts making choices instead of collecting possibilities.

I still use market categories. I just do not let one win a roadmap argument by itself anymore. If I cannot name the person, the moment, the workaround, and the cost, I have chosen where to keep looking. I have not chosen what deserves a month of building.