Why Product Metrics Look Right in Systems That Feel Wrong

Product metrics often tell a reassuring story - adoption is strong, usage is growing, retention looks healthy. And yet, when you look closer, the system itself feels off. Work gets duplicated, users rely on side channels, and teams spend time fixing things that “should have just worked.”

In this session, I’ll explore a pattern I’ve seen while building in startup environments: metrics tend to reflect the part of the product that initiates work, not the part that has to complete it. As a result, we optimise for what’s visible on dashboards, while missing the operational friction happening underneath.

We’ll unpack how this “metrics comfort” emerges, why products in multi-user systems often feel broken despite looking successful, and how to start identifying where work is leaking outside your product. This session will be focused on real examples and practical ways to rethink what we measure - so that our metrics reflect outcomes, not just activity.