Cohort analysis for understanding your customer retention and churn…
Cohort analysis can be applied to a range of business problems. It provides a visual representation of data that quickly highlights performance. However, the majority of businesses apply the technique to monitor and measure retention and churn. Improving retention and reducing churn is vital for businesses to succeed.
Retaining customers of course improves your customer lifetime value. Engaged and loyal customers provide testimonials, referrals and become your evangelists. These customers are good for business.
On the flip side, customers that have a high cost of acquisition and low value to your business – perhaps their average order value (AOV) is less than your average, or they purchase infrequently – are also worth identifying. Knowing these customers opens up opportunities for targeting them with campaigns to increase their value or manage their churn risk.
Identifying such customers is possible through cohort analysis. It provides a visual snapshot to recognise the cohorts of customers that are performing well for your business (or not!)
This analytical technique can help answer questions and provide a starting point for investigating how to:
- Improve customer retention
- Reduce churn
- Increase revenue
With this report you can spot trends in user behaviour, informing your customer understanding and aiding analysis of your marketing activity.
So, what exactly is cohort analysis?
Firstly, what’s a cohort?
Essentially a cohort is a group of individuals who share the same characteristics. Cohorts are subsets of your audience, segmented by their acquisition date – perhaps the first time they visited your website or app, the day on which they subscribed to your newsletter, or the month in which they first made a purchase.
Typically, cohort analysis is an analytical technique applied to better understand a group of individuals – a cohort – who joined, signed up or purchased at the same time.
In practice, a cohort analysis is a visual table that displays cohort data to allow for the comparison between cohorts. For example, comparing customers acquired in March vs February by their engagement or average order value.
Cohort analysis is a standard way of viewing business performance and provides a fast visual assessment of the business direction.
What’s the benefit of adopting this technique?
Put simply, a cohort analysis table provides a clear visual representation of your best (and worst) performing cohorts. The visual snapshot can be used to drill down into problems that need addressing and highlighting successes that should be considered for optimisation.
Comparing customers by acquisition date allows you to perform further analysis into what has driven high performing cohorts. Context is key in order to fully understand the how and why of individual cohort performance, and this requires further analysis. However, without the initial first step of undertaking a cohort analysis, it is impossible to tell which cohorts should be analysed further.
Is it suitable for my business?
If you want to know more about the performance of your customers, then cohort analysis is a great analytical technique to employ.
Cohort analysis is applicable to both B2B and B2C businesses. However, it is vital in subscription businesses.
The visual display of data enables analysts to compare groups of subscribers, users or customers at the same stage of their lifecycle to understand trends. This data can be enhanced and contextualised with marketing campaign data or calendar events, which may have impacted customer behaviour.
Difficulties in achieving cohort analysis
Performing cohort analysis and ensuring the report is useful to the business is not always easy.
In order to ensure accurate and analytically useful cohort analysis, there are some initial steps which must be taken:
- Bringing together all relevant data
- Cleaning the data
All of the relevant data must first be brought together and made available. Typically this includes all data concerning revenue, orders and customers. Sources such as internal databases (SQL servers), point of sale (Shopify), marketing tools (Facebook), data analytics (GA4) and your CRM may all be required. This ensures that all attributes are available for inclusion in analysis. However, often data is stored in different formats and may be duplicated across an array of sources, which can cause problems.
With a data warehouse in place, all relevant and appropriate data can be fed into the analysis and building a cohort table. Then, once the data is ready, the analysis can begin.
However, knowing how to interpret the analysis and use it to explore your business performance is also integral to success. Cohorts are not intuitive and can take a little getting used to as data is displayed in a fairly unique way, for example, triangles and comparative decay curves.
Finally, cohorts should be used to dynamically explore your business performance, however this can be difficult to achieve. Often dashboards are built then simply viewed as a static report, which is of less value to the business.
How to use cohort analysis to drive value in your business
Simple cohort analysis helps determine which customers joined or purchased at a set time. The technique helps to derive actionable insights – the behaviours of each helps you to understand the reasoning behind their performance.
Cohort analysis allows you to ask more specific, targeted questions. Seeing how a cohort performed provides the opportunity to drill down into the why. As a result, you can take actions to improve your acquisition and retention, whilst reducing churn. Use this information to make operational decisions.
This type of analysis is particularly useful when:
- Analysing marketing efforts, for example measuring marketing campaigns
- Understanding trends in customer spending behaviours
- Reviewing website performance
- Looking to reduce early customer churn
Layering acquisition views (marketing channels such as e-mail, social) as a dimension into your cohort analysis allows you to compare various channels and marketing or sales activity, measuring their impact on cohort performance.
Cohort analysis tools for marketing professionals, provided by services such as Google Analytics, generate a cohort table, populated with website data. However, for more advanced cohort analysis, data sources, such as those mentioned above must be available. This allows you to see cohorts not only based on website traffic but add advertising campaign data for example, to understand the reason behind the cohort performance.
How to achieve cohort analysis
The steps typically involved in the analysis process include:
- Pulling all of your data into one place – the data warehouse
- Cleaning and transforming your data to make it analytically useful – grouping all data held on customer for instance
- Calculating lifecycle stages of your customer
- Creating tables and graphs to build your report
A visualisation tool is usually used to build a cohort analysis. This enables the identification of trends and patterns, typically differentiated by colour shades.
Leveraging cohort analysis
Ultimately, cohort analysis enables business users to spot trends and uncover opportunities for further analysis. Moreover, the analysis should generate actionable insights, resulting in change. This may be for example; making alterations to the user journey, targeting offers to a particular cohort, initiating a nurture sequence or activating a loyalty programme.
Whilst tools such as Google Analytics can provide a basic cohort analysis, in order to achieve a more thorough understanding, all business data must be considered.
Leveraging cohort analysis for your business isn’t always easy. But, kleene can make it simple…
Achieving a single customer view can be a difficult task for many businesses. But, kleene makes it easy…
We have worked with customers in a range of industries and sectors from education to ecommerce to provide a clear view of their retention and churn rate. As a result, kleene clients have been able to understand their revenue and retention rate per cohort and enhance this with their marketing campaign data.
Our software makes accessing, centralising and cleaning your data easy and fast, enabling the application of cohort analysis. What’s more, our data team can even build it for you. Need some help with the analysis? Our analysts are on hand to uncover actionable insights!
Read our customer case studies to discover how we’re helping businesses use their data to make decisions.