Optimizing for revenues when A/B Testing Purchases Vs Subscription

[Hugo Hardel]

We have an iOS app that we are optimizing for revenues. We want to A/B test lifetime purchase Vs subscription revenues.

I looked everywhere for information on how firebase a/b testing does the accounting when comparing front-loaded revenues (purchase), vs back-loaded revenues (subscription), but could not find any information. 

My main question is this: does firebase a/b testing account for projected subscription revenue over a lifetime (LTV), or does it account for subscription revenues as accrued?

Our initial experimentation would suggest that a/b tests account for subscription revenues as accrued, which puts subscriptions at a substantial disadvantage (assuming a $10 lifetime price vs a $1/month subscription it would take the subscriptions upgrade at least 10 months to catch up)

Thanks!