I downloaded the raw RevenueCat CSV exports for 2026 from my Google Cloud Storage bucket and wrote a custom script to parse all unique transactions year-to-date, filtering exactly by ownership_type.
I discovered that family sharing might be causing massive distortion in my RevenueCat charts.
In 2026, I had 100 unique iOS users register a transaction in RevenueCat. Here is how they break down:
- Direct Purchasers: 59 users (59%)
- Family Members: 41 users (41%)
A staggering 41% of my iOS "subscribers" are Family Sharing members. These users generate exactly $0.00 in revenue, but if RevenueCat treats them as active subscribers in the charts and if the primary purchaser cancels, then the family members eventually churn out as well, generating multiple "churn" events for what was technically only one real lost customer.
What led me to dig into this were the notifications I got on cancelled subs via the Revenuecat app on my phone and pc. I get bombarded with 5 notifications for a single user subscription cancellation and it made me think that perhaps that’s what is causing the disparity between my android and ios data I see in the benchmarks and analytics tab.
I’m using Flutter so it’s the exact same app on both platforms. I’m also using the same store listings/product pages so I don’t think it comes down to expectations.
I’d like to know if I’m wrong about this and why I would be. Thanks.
