Matching and dating

Safety is the architecture, not a feature list. Photo verification that establishes a person matches their pictures, blocking that is immediate and total, reporting that reaches a human quickly, and a deliberate decision about what location precision is exposed — showing distance in kilometres rather than a position, and never a live one. These are not additions to a working product; a matching app without them is not finished, and shipping one is a decision with consequences that fall on users rather than on the company.

The market's structural problem is a lopsided ratio, and it distorts every metric. In most markets one side vastly outnumbers the other, which means the crowded side experiences the app as sending messages into a void while the scarce side experiences it as harassment. Design has to intervene: limiting how many conversations can be opened, requiring some investment before a message can be sent, and ranking so that attention is distributed rather than concentrated on a handful of profiles.

Then there is the incentive that shapes the whole category: a dating product succeeds commercially when people keep subscribing and succeeds for its users when they leave. Every design choice sits somewhere on that tension, and the ones that optimise purely for engagement — endless swiping, artificial scarcity, paid visibility — are visible to users, who talk about them. Being deliberate about which side you land on is a product strategy question, and it is worth answering out loud rather than by default.