Samaipata’s 3-step/5' cohort analysis (Part 2):
Run a full stack cohort analysis from scratch in just 5 mins and leave your tech team to focus on the hard stuff
By Álvaro González San Pedro
Welcome to Samaipata Ventures’ blog, a VC fund specializing in market-place and disruptive e-commerce models in Europe. If you don’t want to miss out on any quality information, follow us here on Medium, Twitter or suscribe to our newsletter.
If you’ve never heard about cohorts before, we recommend checking out our introductory post before. If you’re already familiar with cohorts and you want further insight, keep on reading. You’ll find our internal cohort analysis template at the end, for you to download and update with your business’ data.
If you’re reading this post it means you already know what cohorts are and what they mean, you know how to calculate your client’s lifetime value (LTV) and you’ve seen and understand cohort analyses. Perhaps what you don’t know yet is how to run the cohort analysis by yourself, without having to rely on your tech team. Have 5 minutes? Yes, it’s that easy, and that quick. If you’re intrigued keep on reading.
At Samaipata we want to share our internal cohort study with you, an easy 3-step Excel template that will allow you to run a quick cohort analysis on your own. The template is not just extremely easy to use (you just need to insert 3 rows of your business’ historical data); it also goes a bit further than traditional cohort studies you might have worked with before.
But first, why is it important to go deeper into cohorts? The answer is simple: average values from a typical cohort analysis can hide many things and make the story seem simpler than it actually is. Think about it, if you have 10 clients and 100 orders for that group of clients, the traditional cohort analysis will tell you that, in average, a customer makes 10 orders. But behind this conclusion there are hundreds of possible scenarios that are unfortunately left behind, but are nonetheless essential to understand in detail how your clients are behaving. Maybe just a few users stay, but become very loyal customers. Perhaps, many stay but are not too active. As you see, both scenarios are radically different.
That’s why at Samaipata we like to separate the recurrence analysis in its two main variables. On one hand, we analyse user retention (that is, how many FTB remain as users past the first month); on the other, the quantity ordered by the actual buyers (how many orders an actual buyer makes in the subsequent months). In the previous example, we could have two extreme cases before us. On one side, 1 single client could have made all of the 100 orders; on the other, each client could have made 10 orders. In both cases, the result is the same: in average, each client makes 10 orders. But as we can now see, behind this “10 average orders per new client” there are two very different possible scenarios. And –important to point out– two scenarios that, in many cases, will require unique and diverging business and marketing strategies.
In the first case, you could be having a problem retaining users. The retention rate is only 10%; perhaps you’ve acquired the wrong kind of customers. However, the users that have stayed seem to be hooked to what you offer (each of them makes 100 orders). In the second scenario, you’re great at retaining customers (you manage to retain 100% of them, thanks to a very targeted and intelligent acquisition strategy), but the “quantity” bought per each actual buyer indicates they are not that “hooked” (each returning buyer only makes 10 orders vs. 100 orders of the first example).
There is no better or worse scenario, especially given that some users are more costly to acquire than others. But of course, the decisions you will have to take in each are completely different, no doubt about that. The situations are very different but at a first glance you could, mistakenly, think that both scenarios are the same.
At Samaipata we like to work in the following way. We first run the traditional cohort analysis, to obtain the average orders by first time buyers (FTB). On top of this, we look for average retention and average quantity bought by actual customers. In the end it’s just disaggregating the analysis into two parts. If the average orders by FTB can be achieved by multiplying R (avg retention rate of FTB) x Q (avg orders per actual buyers), why not obtain each separately and analyze each of the variables in more depth. Are you targeting the right customers, are you retaining them with success? Or at hooking them to your product, to a point where they can’t live without it? Or even better, are you succeeding at both?
In any case, in order to optimize your strategy, the more you know about how your customer behaves, clearly the better. That is why we are sharing our internal cohort analysis template with you, so you can modify it with your business’ numbers and really find out what is going on. Don’t worry, the template is as intuitive as possible and you’ll just need 5 minutes to complete the 3 steps by yourself. Stop bothering your tech team with things you can do on your own and let them focus on the hard stuff!
What are you waiting for to try it out? If you have any doubts with interpreting or updating the data, we’ll be delighted to help out. Email us!
Click here to download the template. If you have any problems downloading please contact us.
For more quality analysis follow us on Medium.