Solved

Experiment subscription vs. consumable based paywalls

  • 20 January 2023
  • 1 reply
  • 113 views

Badge +2

I’d like to test if my app works better with subscription based or consumable based model.

In subscription-based model, users will be presented a Paywall Screen to subscribe weekly to be a “premium” user, so they can play unlimited games.

In consumable-based model, users will be presented a Paywall Screen to purchase X number of games. When they play X number of games, they will be presented another Paywall to top-up.

Is RevenueCat Experiments suitable for this use case? 

 

Couple of points I’m curious are:

 

  1. Does Experiment Results work as expected with different monetisation models? Especially LTV is the concern point.
  2. With the nature of the Experiment, I need to show “Remaining Games”, “How To Get Refund”, “Current Subscription” etc. information in different screens. This means I need to know the bucketing value of the user. Does RevenueCat Experiments provide this? (I guess I can check current offering in different screens to get the experiment bucketing value and show correct UX)

Thank you

 

 

icon

Best answer by Michael Fogel 25 January 2023, 22:57

View original

1 reply

Userlevel 3
Badge +6

Hey @berk ! 

 

As long as those two product setups are represented by distinct offerings, and the products that get served to customers are controlled by our getOfferings endpoint, then yes the experiment will work!

 

I would just recommend that reviewing our Results Docs first to make sure that the metrics that you’d be looking to track to measure success are available. One specific case that may be problematic for you, is that we don’t count the number of purchases made by a customer anywhere. However, we count the number of customers making at least one purchase, and the total revenue, so you could use that to back into the average number of purchases per customer if you know the price of your consumables — but you’d have to do that math on your own as of now.

 

I hope this helps! 

Reply