The Ultimate Guide to Cohort Analysis for SaaS

When it comes to fighting churn, cohort analysis is one thing SaaS businesses must have.
It’s one of the tools you can depend on to find and fix the causes of churn, which is one of the most common reasons that SaaS businesses fail.
This guide will discuss what cohort analysis is, why it’s important for your SaaS business, and how it can be used to grow your product by combating churn and increasing retention.
TL;DR
- Cohort analysis is a type of behavioral analytics that shows you what a sub-section of your user base (also known as a cohort) is doing in your SaaS product.
- It is important because SaaS businesses rely on customers to renew their subscriptions in order to grow, and cohort analysis arms you with insights needed to keep them from leaving too early.
- There are two types of cohort analysis: acquisition cohort and behavioral cohort. Acquisition cohort groups a set of users by when they joined your product and behavioral cohort lets you see the events that lead to their churning.
- Cohort analysis is conducted using a cohort table.
- Some cohort analysis tools include Google Analytics, Barametrics, and ProfitWell.
- To use cohort analysis to fight churn, you need to define clearly what you plan to achieve, explore the cohort analysis data, create a hypothesis based on that data insight, implement and test your hypothesis, and then review the results.
What is cohort analysis?
In summary: cohort analysis is a type of behavioral analytics that helps you see what a sub-section of your users (a “cohort”) is doing in your tool.
The metric you’re usually measuring with cohort analysis for SaaS is user churn – the moment when some users stop using your product and leave. That’s why you can also call it “customer churn analysis”.
The two most common types of cohort analysis are:
- Acquisition cohort: this captures a user segment based on the time they signed up for your product.
- Behavioral cohort: this one categorizes a cohort based on their behavior while using your product.
Depending on the one you conduct, cohort analytics can get you answers to the following questions:
- For how long does an average user stick to your product and at what point do they churn?
- Which segment of your users (e.g. by persona, by plan, etc.) churns the most?
Cohort analysis is done using a cohort table:
Here is the breakdown of everything in the table above:
- A row (from top to bottom) segments a cohort by their signup dates (particularly by months).
- A column (from left to right) indicates the amount of time that has passed since a user subscribed to your product.
- Each cell shows you the percentage of users that churned within a specific month of subscribing.
Now that we know what cohort analysis is – let’s see why it’s important!
Why is cohort analysis important for SaaS?
A successful SaaS product has one important characteristic: stickiness.
If your product users are falling off your radar before reaching their maximum lifetime value (CLV), you need to know why.
By looking at when the users churn and how the churning rate changes based on when a new cohort signs up for your product, you can see if your product and marketing strategies are working or not.
Cohort analysis will help you to do the following:
- Understand how long it takes for an average user of your product to become disengaged and leave.
- See if your product is sticky or used as a one-off.
- Learn if the changes you are making on your new user onboarding overtime is impacting your first-month retention rates.
- See if the change you made to your product (e.g. new products, new products line, new offering), your UX, or your marketing strategy (e.g. promo, new discount rate, entering a new channel) had any impact – positive or negative -on your product.
- Improve your retention goals.
Watch the video below to learn more.
In summary, the insights you gain from cohort analysis can be useful in creating an actionable retention strategy e.g. targeting the ‘sensitive’ months in your user journey.
What types of cohort analysis can you use?
Acquisition cohorts
Acquisition cohorts group a set of users by when they joined your product. The unifying factor here can’t be any other thing than their signing up moment (the month).
An acquisition cohort can show you:
- How the changes you make to your product over time impact a whole set of users.
- If your product strategy is heading in the right direction (reducing time-to-value, increasing stickiness + retention)
But there is a limit to what the acquisition cohort can do for your SaaS businesses on its own, especially when it comes to getting a deeper understanding of why exactly your customers are churning.
Behavioral cohorts
A behavioral cohort is where you dig deeper into the events that lead to churn.
This lets you segment users by:
- The plan they subscribed to
- What channels the users were acquired through
- Which features they engaged in your app
- What actions the users have taken
- User persona (e.g. role, company size)
As you can see by now, the number of insights you gain through this analysis is incredibly useful, and can help you with:
- Understanding which set of users get the most value from your product.
- Making informed marketing strategy decisions (what types of personas should you target? Which channel brings in the most qualified leads?).
- Understanding which plan brings the best value for your money.
- Understanding which features of your product you should push users to adopt – since you can see that the use of these features correlates with retention – through in-app experiences, onboarding, and personalized calls.
How to read a cohort analysis table
At first glance, a cohort chart can appear very intimidating. But in reality, they are very easy to read once you understand the chart’s components.
Using one cohort, here is an overview of a cohort chart in Baremetrics:
Now, let’s deconstruct this.
From left to right, here is what you are looking at:
- April 2019: This is the ‘cohort’, as it stands for the month the users signed up.
- 24: This is the number of customers in this cohort. It means that 24 customers signed up for your product in April 2019.
- 92%: This represents the percentage of customers remaining in the first month.
Every other column you see after that shows you the percentage of customers that remained within that cohort after each month. For example, under the “1” column, 88% percent of the original signups (the April 2019 cohort) remain after their first month.
And in their second month, only 79% of those 24 remained – and so on.
You can also do away with the percentages and look at the data in absolute numbers. Instead of seeing the percentage, this shows you the exact numbers of remaining customers after each month:
If it’s your revenue you want to see instead of users, you can also use cohorts for that:
But that’s not quite all yet – the real value in studying a cohort analysis is what you do with it, which we will discuss shortly.
Best cohort analytics tools for SaaS
Google Analytics
A Google cohort analysis report allows you to isolate a cohort based on shared characteristics and analyze their behavior.
The isolation can be based on:
- Cohort size: This is selected on the basis of the number of users you acquired in a particular month. For instance, if you acquire a group of users in April 2019 and another in June, and another in July, Google cohort analysis lets you select which of those groups of users you want to examine.
- Metric: Google analytics also allows you to isolate users based on what metric you want to analyze about them. The default metric is user retention, which is how you measure the number/percentage of your users that are returning each month.
- Date range: Setting the date range will enable you to categorize users by the window of time you want to study, e.g. the last three months.
- Cohort type: The only option here is acquisition time, which indicates what time a particular set of users first interacted with your product.
You can use Google cohort analysis for both your website and product, and the best part is that it’s free!
To access Google cohort analysis, sign in to your Google Analytics account, then go to Audience, and choose Cohort Analysis.
ProfitWell
Aside from providing you with cohort analytics, what sets ProfitWell apart from other cohort analytics tools is its advanced algorithmic features, such as:
- Customer recovery and churn reduction: Unlike in Google Analytics, where you have to make sense of your data by yourself, ProfitWell provides you with insights on what exactly is causing churn even if it’s unrelated to your product.
- Pricing optimization: ProfitWell also shows you real-world data about your industry and its market. It helps you eliminate the guesswork about pricing so you can set the best prices for your business.
- Revenue recognition: ProfitWell gives you deeper insights into your revenue sources by showing you the exact source of revenue and fluctuations such as upgrades or downgrades.
While its cohort analytics tool comes completely free of charge, ProfitWell’s premium service starts at $1,000/month.
Baremetrics
Baremetrics is an analytics tool for SaaS and subscription businesses.
It shares all other characteristics with other cohort analysis tools but stands out with its ability to integrate with Stripe and track additional metrics like recurring revenue, churn, and average revenue per user.
The basis of Baremetrics’ design is from the idea of pulling data directly from Stripe and aggregating it on a single dashboard. However, despite this innovation, Baremetrics doesn’t have as many features as ProfitWell or and is not a permanently free analytics tool like the previous two examples.
The pricing for Baremetrics begins at $350/month, and includes the following additional features:
- Recover, which helps you recover lost payment as a result of failed payments.
- Cancellation Insights, a feature that shows you why customers are canceling and sends automated emails to bring them back.
How to use cohort analysis to fight churn in SaaS
Step 1: Define your goal
What do you want to achieve and within what timeframe?
For example, if you own a social media scheduling tool for agencies, a clearly defined goal could be to increase the number of your ‘Pro’ plan subscribers within a particular month.
Step 2: Explore your data
Go into your cohort analytics tool and see the current retention rate in a behavioral cohort is.
This will tell you if the common traits of a selected cohort are positively or negatively impacting your retention rate.
Step 3: Create a hypothesis
Create a hypothesis about what could improve your retention rate in relation to the outcome of your data exploration.
For instance, in the case of a social media scheduling tool for agencies, instead of just using onboarding experience to point users to the two key activation points (adding social media accounts and scheduling their first post), you can also add some key productivity features such as:
- Where to add their team members
- How to group accounts (e.g. by clients)
Step 4: Implement and test your hypothesis
Having developed a hypothesis, implement it to see if you are right or wrong.
For example, you can use Userpilot to build a new onboarding experience. It is code-free and can be deployed in minutes.
Step 5: Review the results
By building a new experience for your users based on your hypothesis of what works, you can now review the results to tweak and adjust any future plans.
Conclusion
Cohort analysis is a simple task that sounds complex. We hope we have been able to simplify the process of conducting cohort analysis for your SaaS business through this post.
SaaS businesses rely on customers renewing their subscriptions. If they aren’t, then your business will not grow. But with cohort analysis, you will see the why behind your customer churn, which will let you understand the reasons in time to do something about it.
If you want to learn more about how to maximize your product growth and improve your analysis skills, get in touch to schedule a free demo with Userpilot!