Retain - Retention. Payment. Security

I've shipped 76 browser games. Games are built around one question, what makes someone come back tomorrow, and after enough years of that you start noticing it everywhere else. I also work in behaviour change professionally, where the same problem appears at population scale: knowing what to do is not what changes what people do. Habit, timing and environment decide that. Then I looked at my own analytics and found week one retention sitting at zero across every cohort. I had built the wrong thing well. That was the point where retention stopped being an interesting problem and became the only one worth working on. The problem Businesses spend to acquire customers, then lose most of them quietly. Roughly 77% of users are gone within three days, and the median consumer app keeps 4 to 7% past a month. The tools that exist either report what already happened, or send messages and take credit for whatever the chart does next. None of them can tell you whether the customer would have come back anyway. What changed with the launch Two things, both from actual conversations rather than planning. Every business owner I spoke to could tell me their signup numbers. Not one could tell me their day 7 retention. The product I thought I was selling, better interventions, was not the problem. Not knowing was the problem. Two of the four also flinched at the word notifications. Both were building in sensitive categories, and both wanted the measurement without anything reaching their customers. So measurement became the entry point, and the intervention became a switch you turn on later rather than the thing you start with. The third change was to pricing. If the whole claim is that we can prove what we caused, then charging for anything we cannot prove is incoherent. So the holdout is free and permanent, and the invoice reflects the measured difference, or there is nothing to invoice.