Calculating subscription LTV

  • 27 July 2021
  • 4 replies
  • 70 views

Badge

Is there a way to calculate LTV easily?

Can you add an LTV parameter in the dashboard?


4 replies

Userlevel 4
Badge +8

Is there a way to calculate LTV easily?

Not really an “easy” way with subscriptions. Even a $0.99 per year subscription could make you a billionaire if the subscription lasts long enough 🙂 Usually a more interesting metric are 3-month, 6-month, 12-month customer values - to estimate payback periods and such.

You can read more about the LTV calculation we’ve been playing around with in our A/B testing tool here: https://www.revenuecat.com/blog/price-testing-for-mobile-apps

 

A simpler metric to calculate is total spend, or average revenue per customer. This can quickly be seen in the Customer Lists section of RevenueCat. You could create some custom lists here to see the average revenue per customer (ARPU) and average revenue per paying customer (ARPPU) of more specific cohorts too.

 

Can you add an LTV parameter in the dashboard?

Once the LTV model powering Experiments get’s a little more fine-tuned it will probably get applied to more charts and other areas throughout the dashboard.

 

I updated this from a question to a discussion because I’m curious to hear if any other developers have rolled their own LTV calculations.

Badge

Thank you Ryan, I’m looking forward for that feature.

 

Michael.

I’m looking forward for that feature too!

I agree with Ryan in that the “L” (lifetime) of LTV makes calculation a loaded topic. The true lifetime of a customer is at worst impossible and at best irrelevant in most situations.

 

Most app developers pick a timeframe (180 days / 1-year) and model expected revenue at that point. And while the most advanced have built custom ML-models it’s overkill for most. 

 

A simple regression curve works for most: some examples.

 

I will caution against relying too heavily on ARPU for projections -- we’ve seen quite a few missed opportunities by leaning on averages.

 

Reply