Share this content

My key KPI: Customer cohorts

20th Feb 2019
Share this content
My key KPI

My key KPI is a recurring content series where we ask CFOs and FDs what metrics and measures they use to drive their businesses forward.

The aim is to understand how different finance professionals across a broad array of industries and sectors use data to inform their decision making.

In this latest edition, we speak to Joe Gallard, co-founder of Reducer and former FD at Gousto and Carwow.


My key KPI: Customer cohorts

What if you were actually underestimating the lifetime value (or LTV, if you want to get all technical) of your customers? The dangers of overspending on customer acquisition are well noted, but the cost of not squeezing the full value from each customer is often overlooked.

“Normally with an LTV calculation, you underestimate the long tail value of customers you acquire,” said Gallard. “It means you don’t invest as much as you should.”

For example, Gallard points to his previous work with Carwow, a platform for buying new cars that uses a reverse marketplace model (ie, where buyers post ads of what they want to buy, rather than sellers posting what they sell).

Through a customer cohort analysis, it became clear that people weren’t just buying one car from Carwow. “People were coming back and buying third, fourth cars from us. It changed the value you ascribe to someone.”

So what is a customer cohort? Simply put, rather than looking at all users as one unit, a customer cohort bunches users into related groups. Gallard uses months. That is, the month the cohort was acquired.

These cohorts usually share common characteristics. So the January cohort’s behaviour over the last six months can inform your strategy for the July cohort, for example. Patterns manifest over time and the business can tailor it’s marketing spend to match that.

To return to the Carwow example: logic may have said to taper off with marketing spend on a customer that has made a purchase -- but cohort analysis proved otherwise and subsequently, the company was able to squeeze more value from each user.

This all feeds into a calculation that many investors are intensely interested in, said Gallard. LTV/CPA takes, essentially, ‘how much you get from someone’ and divides it by ‘how much it costs to get them’. A favourable LTV/CPA is the driving force behind Netflix and Spotify’s titanic valuations.  

“You need a database that tracks all this info,” explained Gallard. “In a tech business, it’s fairly simple. You use a SQL table tracking users and purchase history. And then I map it out in Excel based. I only use Excel. It gives the flexibility to do things in your own way.”

Replies (0)

Please login or register to join the discussion.

There are currently no replies, be the first to post a reply.