How ARPU can lead to a 120% error in Customer Lifetime Value

blog.custora.com

38 pointspospischil6 comments

I've done many, many LTV and CLV reports for companies and I always account for this. Protip: churn also goes down for SaaS apps. Also, you need to account for a discount rate but most of the time people don't want to hear this because it makes their numbers look less shiny. Pretty basic math, all things considered.

Churn also goes down in retail. If you look at retail cohort analysis, you will say that 10% of users are active three months after their first transaction, and then 5% are still active six months later. So you lose 90% in the first three months, and then only 50% in the next three months.

This is particularly important in the RLV calculations, as assuming a constant churn rate will undervalue your existing customer base.

Sounds like you are taking the right factors into account. What sort of accuracy are you getting with your calculations?

Pretty close. Usually around 5 to 15 percent when taking off the top 2% of customers.

I suppose fitting a decaying function would be better than flat-out ARPU?

Or even a moving weighted average?

We're going to explore some of those approaches in upcoming posts!

Whatever font you guys are using for the body of the post is very difficult to read.

Reply on news.ycombinator.com