RevenueCat · B2B · in-app subscription software
Start on Amplitude; add a data pipeline once product changes break your definitions
Co-host says it smoothed definition changes for data users; no numbers
Workedanalytics-attribution
What they did
Started with off-the-shelf Amplitude plus a few semantic events (app created, project created) and a generic page-loaded event with the path broken down, which allowed retrofitting later questions. When a restructuring of the product changed what core terms meant, added an ETL step that encodes each definition change by date into a normalized 'super definition'. Now close to running its own event tracking on top of its data warehouse.
What happened
Co-host says the pipeline took the sharp edges off definition changes that were hard to handle in a third-party tool; not needed at launch.
In their words
Sub Club by RevenueCat · Building an Effective Data Product Strategy — Taylor Wells, News Corp“And that's where, for us, it made sense to introduce some sort of pipelining step, cuz then what we can do is encode that change into a pipeline and be like, "Okay, like, in this month of this year, we changed this definition,"”
Related topics
Get tactics like this every Monday
New tactics from the week's founder interviews, each linked to where it was said.
More from RevenueCat
All 9 RevenueCat tactics- Launch a mobile app with a home screen people want to screenshotAbout 100x more sightings on X overnight; mobile usage near dashboard usage
- Email batchmates, then install your SDK in their app yourself at a coffee shopLanded early customers like Podcast App; no numbers
- Ask users how they heard about you, with a free-text boxRevealed ChatGPT and Claude were sending people; read daily
- Build the product so free users cost almost nothing to supportLets the company keep a generous free tier; no numbers
- Check 5-10 key numbers every morning, with 30- and 90-day averages30-day above 90-day means healthy; no outcome reported
- Publish a yearly benchmark report built from your customers' anonymized data
More from this episode
- Disney+: Build your own event tracking, keep Adobe as a backup, then drop Adobe~$1K/day in-house vs ~$33K/day for Adobe; Adobe removed within weeks
- Disney+: Log one generic 'button clicked' event with keys, not a new event per featureGuest says it avoided costly renames and merges; still in use years later
- Disney+: Email the boss three bullet points by noon each day instead of dashboardsThe president stopped needing dashboards; no numbers
- Disney+: Move a show with near-100% completion up from the last row of the home screen
- Disney+: Group viewers by household, like parents or single adults, not by fan typeHousehold groups predicted behavior better; over half the audience were single adults
- Dump app event logs into a plain SQL database instead of buying analyticsToo hard to explore, so nobody asked questions; moved back to Mixpanel
- Buy analytics tools first, but set dates to decide whether to switchGuest says companies without checkpoints get stuck with rising vendor costs; no numbers