+1 vote
We are checking out ChartMogul, while their dashboard looks good, we find that the numbers and calculations do not stay consistent. Also, their LTV treats annual subscriptions and monthly subscriptions the same, which I do not believe is correct. What do you use? If you use Chartmogul, are their calculation and numbers to be trusted?
by (340 points)
What do you mean they treat them as the same?
To calculate LTV, they convert annual subscriptions to the monthly equivalent (divide by 12), then they apply the same churn rate to all subscriptions. I think this is incorrect. There should be different churn rates applied to the annual and monthly, since annuals are not up for renewal at the same time increments.
LTV should just be the cumulative sum of all the monthly payments grouped by the start date of the subscription. Do you have a screenshot that shows this?
The model that chartmogul uses is MRR/[Monthly Churn Rate] or more simply subscription price* lifetime. So a $10/m subscription and a lifetime of 4 months equates to a $40 LTV.

1 Answer

+1 vote

I can share the approach we use to calculate subscription LTV:

Lifetime

We calculate a max lifetime for each cohort:

  1. For cohorts with 2-5 renewal events, project the rest of the profile based on historical averages by subscription and country, to project maximum lifetime.
  2. For cohorts with 6+, use the actual retention profile and fit curve to project maximum lifetime.

Value

Use avg. subscription price, weighted by duration and country, but normalized to a daily value.

LTV = Lifetime(days) * Value(daily)

A limitation related to one of your other threads is around treatment of annual subscriptions, oftern with no renewal data, for newer apps or newly launched annual plans. I'm curious how others are treating this.

by (160 points)