Surprising data analytics examples that drove unexpected product decisions

Been digging through old campaign data and found some weird patterns that completely changed our roadmap.

Curious what counterintuitive insights others have discovered that made you pivot hard on product or growth strategy.

Push notifications were hurting our retention. We thought reducing frequency would help - it didn’t. Users receiving 3 or more daily pushes had 40% higher lifetime value than those getting just 1 or 2. The issue was not volume but timing and relevance. Low-frequency users received random alerts, while high-frequency users got tailored messages based on their behavior. We switched to triggered notifications and achieved 25% better retention with 60% fewer complaints. Sometimes you need to increase your efforts, not reduce them.

Heavy users were actually churning faster than casual ones. They would burn through content too quickly and get bored. We throttled features for power users and added some friction to slow them down. Retention jumped 30% because people stayed engaged longer instead of burning out.

Free users converted better than trial users somehow.

Our worst ad became our biggest winner after we switched up the landing page.

Everyone wanted to axe this boring ad with a 0.8% CTR. But those users had 3x higher LTV than our flashy creative pulling 2.1% CTR.

The plain ad attracted people who actually wanted our product - not just tire kickers. We kept it running and built a custom funnel for that traffic.

Sometimes you gotta double down on what looks like a total failure.

Ads that seem to flop can sometimes deliver high-value users. Low-cost clicks might not convert well but can lead to longer retention.