Ashley Stirrup

Your metrics are green. Your customers are leaving. Ron Kohavi calls these "watermelon metrics." Green on the outside. Red on the inside. Here's how it happens. A team picks metrics that feel right: pages visited, features used, support tickets resolved. These go up. The team reports success. Executives see green dashboards. Meanwhile, customers are quietly leaving because the product is frustrating, slow, or missing what they actually need. The metrics didn't capture it. The problem isn't laziness. It's a deeper issue: measuring what's easy instead of what matters. Revenue goes up — but because you raised prices, not because customers love you more. Engagement goes up — but because your app is confusing and people are searching for what they need. Support tickets resolved go up — but because your product is generating more support tickets in the first place. Each of those is a watermelon metric. The signal looks positive. The underlying reality isn't. Kohavi's fix: the OEC (Overall Evaluation Criterion) should be tied to customer long-term value, not team output or surface-level activity. For most businesses, that means something like: are customers getting the value they came for, repeatedly, over time? That's harder to measure. It requires thinking carefully about leading indicators, not lagging ones. It requires being honest about which metrics can be gamed by your own team. But it's the only way to know if you're actually winning — or just winning on paper. What's the most dangerous watermelon metric you've seen in the wild? @ronnyk

Only one snapshot so far — the engagement-over-time curve appears once the daily scrape has captured this post at least twice.