*** In Cohorts We Trust *** 🐬
Template to build your cohorts included
Cohort analysis is one of the best ways to understand your customers. It’s a fantastic method to observe current customer behavior and predict how they will act in the future. This could include determining when their next order will be, how much they are worth and the potential turnover you will generate in the coming years. Because rumor has it, a well-designed cohort analysis is vital to determining your Customer Lifetime Value (aka CLV). It’s a bit like the Holy Grail: we often talk about it, but it is still hard to grasp.
That being said, it still remains a real hassle to efficiently translate raw data into a proper cohort analysis. As we love data (and are always seeking ways to better understand product usage), Eric and I got pretty passionate about cohort building and analysis. We even built an in-house tool that we’ve started to share with a few entrepreneurs who had yet to really focus on it. In the spirit that this perspective can help others in the ecosystem, we’re excited to open source our template below.
→ For those who cannot wait to get their hands dirty, we’ve attached a link to it at the end of this article ;)
Spoiler alert: we just got hunted on Product Hunt, feel free to have a look -> https://www.producthunt.com/posts/in-cohorts-we-trust
The template we share with you today focuses on consumer models and marketplaces. It has been iteratively built and tested with some of the Serena portfolio companies. Special thanks to Nicolas (Alittlemarket), Aurélien & Théo (Cheerz), Nicolas (Malt) & Alexis (Kima) for reviewing and challenging our draft.
Our aim for you is simple: less time spent on building cohorts, more time spent on segmenting, analyzing and challenging your data. Keeping in mind that no businesses and no databases are the same, you might need to make some adjustments to adapt it .
If you have any feedback or if you just want to talk about cohorts, metrics or marketplace models, feel free to get in touch with us: egossart@serena.vc / sleroy@serena.vc.
Thanks for the model, but why are cohorts so awesome? 🤩
Long story short, a cohort is a group of users or clients who have a common characteristic, oftentimes this is determined by the date of a customer’s first order but not always… (i.e. stay tuned for more on that). It is a way to look at how your customers behave and how these behaviors evolve in the long run.
Analyzing your cohorts allows you to accurately:
1. Understand your customers and how they use your product
2. Anticipate how your new customers will act and predict their behavior
3. Understand your churn according to product evolutions
4. Build your business plan for next year with informed assumptions about your new cohorts based on the evolution of your product
Cohort analysis not only gives you the means to measure how much users really love your product but also how to develop your revenue model: when you acquire a new customer, you know that, on average, he will make a 2nd order X-months later of Y-€ and a 3rdorder Z-months later…
Ultimately, cohorts are a great way to shape your future business model.
Cohorts are not just for CEOs
While cohorts are often thought to be in the monopoly of the CEO, all teams should be involved in this analysis and monitoring as it is an infinitely better way to understand your customers and their actual usage of the product.
Below we’ve added a few illustrative examples to show added values for each team.
There’s no denying that cohorts are crucial for a company at every stage: from its inception in order to understand if your product has achieved product market fit to later down the line to understand how to drastically increase your customer retention and overall revenue.
So how should I segment my cohorts?
Good question. You may have guessed it by now, but how you decide to segment your cohorts is a very important decision. Customers are often grouped according to their 1st order or 1st sign up date. Though this may be a good solution for some of you, we suggest that you don’t limit yourself. The truth is, there are many other ways to segment your cohorts.
It is actually useful to base cohorts off a smaller segment of your users. Here are some examples of the kinds of segmentation that we see most frequently:
• Per type of customer: if you have both SMBs and large corporate customers, it would be surprising that they behaved in the same way and spent the same amounts of money
• Per device: separating your customers who order on desktop vs. those who order on their mobile could be interesting
• Per country: from one country to the next, behavior can drastically vary
• Per acquisition channel: if you have customers coming from inbound and outbound channels, this will probably have an impact on the average basket and repeat. Careful, this is not interchangeable with the above segmentation “per customer type”.
A fortiori, it is worth segmenting your customers according to their usage and purchasing behavior.
Want to give it a try ?
We did our best to make it as easy as saying “Hello”. As for all metrics, we emphasize the importance of not following too many different cohorts. Before you start building your cohorts, it’s always a good idea to spend some time figuring out the right questions to answer. You can get started by using our model and adjusting it to fit your needs.
And of course, once you’ve played around a bit with your cohorts, don’t forget to think about identifying a segmentation that is most inline with your business.
>>> A SIMPLE AND STRAIGHT FORWARD MODEL TO BUILD YOUR COHORTS AND GET RID OF HASSLES <<<
Love,
Eric & Sébastien
***
About Serena
Founded in 2008 by entrepreneurs for entrepreneurs, Serena invests in bold ventures and provides them with an unrivaled level of expertise and operational support.
In just ten years, the firm has already contributed to the emergence of many success stories and has currently 40 handpicked startups in its portfolio such as Dataiku, Evaneos, Malt, iBanFirst, CybelAngel, or Lifen.
Find out more on serena.vc
