Blogs

Using churn data to filter investments

By Waseem Khan, CFA

Head of Research

EDGE Research & Consulting

Posted on: 16 Nov, 2022

When filtering private investments, one key problem is deciding which companies deserve more time; private due diligence requires mobilizing considerable resources and not every effort is worthwhile. As such, initial screening is usually some form of “ghetto due diligence” which relies on a quick and dirty method.

While we have toyed with a number of “ghetto solutions”, one method we use is to simply focus on churn data. The idea is quite straight forward; the basic question we ask when filtering any business is asking whether the product on offer solves any real life problem worth solving. Questions on ability to monetize can come much later.

If the company is indeed adding value for customers, this should be reflected in the way the customer interacts with the product. Stickiness and referrals should be high, and you should see increasing usage as users become more used to the value proposition. In terms of verifiable data, customer churn should be low. A few conversations with existing users can reveal quite a bit of depth in terms of what is going on with the product.

Take bKash for example, monthly transaction value per active user went from BDT~12k to BDT~18k between 2015-21; active users grew 3x during the same period. This can only happen when a service adds increasing value for a user.

Similarly, talent churn can indicate some interesting details. Employees are company insiders, hence have firsthand information on where the firm is going, churn for key employees could suggest they see limited scope for success. On a more basic level, talent churn could indicate lack of incentive alignment.

Given talent has a network effect of sorts (talent attracts talent and vice versa), this can mean the difference between having a team that can execute vs one that just fizzles away. Moreover, conversations with ex-employees can be very useful, having first hand feedback from people who have worked closely with founders is a very handy hack.