"LTV fills in the blank spaces between ARPU and retention, showing you where you stand."
One of the biggest drawbacks of calculating so heavily based on ARPU, and all other traditional LTV calculations, is that there needs to be a large enough sample size (number of customers) in order for the LTV calculation to be meaningful. If a company has a large number of both customers and revenue, then there should be low variation in month to month changes, unless the company is either doing something very right (in which case, we’d expect these numbers to increase drastically) or very wrong (decrease drastically). The problem with having a low number of revenue (or customers) is that it reduces the statistical power of the model, meaning that the model’s expectations are less reliable. Also, with a lower number of customers, each individual customer has a greater impact on the LTV calculation for that month.
Here at ProfitWell, we love to geek out on data and metrics. A lot. Probably more than we’d normally admit in public. After a lot of pondering and discussion, we feel like we've figured it out a better way to deal with low customer numbers and retention by developing a unique algorithm (the technical term being “secret sauce”). This algorithm lets us better regulate small changes, as it looks at how LTV is trending instead. It does this by comparing this month’s LTV to last month’s in order to remove the spikiness that can be caused by small samples sizes, giving an overall more accurate LTV.
Also, by looking at trends instead of using other methods to deal with highly variable data like lifetime capping—saying that each customer will only stay with a company for a certain number of months or years (a number that will be different for each company)—we can also better predict an LTV for all types of companies, regardless of their size, retention numbers, or growth rates. Using trends also helps deal with problems like having months with 100% or greater retention.
Join over 70,000 growth pros from companies like Uber, Pinterest & Twitter