I want to do A/B testing for subscription prices on iOS. The revenuecat docs recommends to “Use distinct Subscription Groups on the App Store for each RevenueCat Offering when testing new prices”.
Here’s what I would then setup:
Price 1 subscription group:
- Annual - $39.99
- Monthly - $7.99
Price 2 subscription group:
- Annual - $19.99
- Monthly - $3.99
I start the experiment, some users are being shown the first subscription group, some others the second. A week lapsed, price 1 group is the clear winner, yay! I stop the experiment. The faq says the following:
When I end an Experiment, what Offering will be served to the customers who were enrolled in that Experiment? | When an Experiment is ended, all customers previously enrolled in it will be served the Default Offering the next time they reach a paywall in your app. |
Therefore everyone is now presented with the same offering that corresponds to the price 1 subscription group on my paywall. Let’s suppose someone from the retired price 2 subscription group wants to upgrade from monthly to annual, they would now be presented price 1 subscription group and instead of upgrading, it would subscribe again, the user now has 2 subscription.
Am I missing something?