Book a strategy call

Analytics, Attribution and A/B Testing

MMP setup and audits, SKAN conversion value models, event design, experiment design and sample size, and building one reconciled revenue view across RevenueCat, product analytics and the ad platforms.

Bad data does not produce cautious decisions, it produces confident wrong ones. These pieces cover how to make the numbers trustworthy before you scale spend against them, and they are written for anyone who has watched three dashboards disagree about the same week and had to defend one of them in a budget review. For the service, see A/B testing and data analysis.

Event Tracking Accuracy: Why the Analytics Events You Trust Are Often Wrong cover image Blog

Monetization

Event Tracking Accuracy: Why the Analytics Events You Trust Are Often Wrong

A trial_started event fires on paywall view instead of trial activation — the chart looks smooth, the number looks plausible, and a quarter of pricing decisions get made against a metric that quietly means something else. This piece explains why event definitions drift, why confidently wrong data is more dangerous than missing data, how the error compounds through experiments and ad algorithms, and the discipline that catches it before it ships another decision.

8 July 2026 · 25 min read

Why Mixpanel, RevenueCat, and Stripe Show Different Numbers — and How to Reconcile Them cover image Blog

Monetization

Why Mixpanel, RevenueCat, and Stripe Show Different Numbers — and How to Reconcile Them

Every month the growth lead, finance, and product open three dashboards and get three different numbers for the same metric — then burn the review arguing about which one is right. This piece explains why Mixpanel, RevenueCat, and Stripe were never built to agree, the four structural mismatches behind the spread, and how to reconcile them by assigning one source of truth per question instead of chasing a single number that doesn't exist.

7 July 2026 · 22 min read