COHORT ANALYSIS FOR A SUBSCRIPTION BASED BUSINESS MODEL
Hi, everyone! Every company wants to keep its customers longer. The term of customer satisfaction is the key to keep customers. We can explain customer satisfaction as an ongoing process. Moreover, during this process, churn rate analysis likens to a camera to take a photo of customer satisfaction at a specific time for some specific groups of customers in the company. The results of the customer satisfaction might reflect as churn or retention on numbers for KPIs. In this article, we will talk about the cohort analysis for a subscription-based business model.
Customer churn is the complement of customer retention. If a company has a %25 retention rate, its churn rate is %75. These rates might change up to industries, companies, or products. After customers bought the product or service, and the customer has been satisfied in his or her first experience, he or she would prefer the product or service again so he or she would retain and would be profitable customers.
Customize your analysis
While analyzing, we can group the customers in terms of specific dimensions. Some groups can be created up to seasons, months, weeks or days. The other option might be for grouping them according to demography. The examples might be extended for grouping methods. It depends on the purpose of the analysis. We call these specific groups as cohorts that give the name of “cohort analysis” to the analysis. In this article, we choose monthly cohorts.
In the company that I am working for, there are subscriptions for 3–6–9 months. It has a subscription-based business model. Subscription renewals increase the lifetime value of a customer. Our customers have a schedule for daily speaking classes during weeks. In the cohort analysis, we explored that we have to repeat the analysis for each type of subscriptions. By the way, not only for cohorts but also, we were able to comment about product life. For example, 3 months service has different characteristics from 9 months subscriptions as well as 3 months subscribed customers are different from others.
Some SQL commands to pick the relevant data;
Before we started the analysis we specified the cohorts and prepare the SQLs to retrieve the data. We defined some SQL variables for the dates and created a while loop in the query. It was a simple query that shows the count of students who have active speaking lessons in monthly periods. Additionally, the cohorts are monthly. For example in the analysis, we have 16 cohorts. In the 16 months, we observed active lessons in the company database for each customer in the cohorts. We used an excel sheet to write the SQL results and visualize the data for a summary.
In Figure-1 above, we blurred the data in the concern of concealed data. We interpret that product life is five months from the analysis. This is our three months subscription type. We provide an opportunity of starting the program in two months. This explains the period of subscription is three months but we add two months more to product life. The latency of the start would be enhanced during special campaigns. Our customers love the latency for the start of their programs. They can adjust their busy schedules until they start the speaking program. We noticed, after 5 months from the start, our customers churned. We can compute the average of all cohorts’ retention rate. This will give a general idea of the retention rate in these cohorts. We can define a range for the churn rate, for example, %x to %y is the retention rate for these cohorts. Then we can look at it closer to find which cohort has the best ratio and, which cohort has the worst ratio. We can think of the marketing strategies, campaigns, or even ads in those months. We can repeat the strategies for the month in which we had the best retention rate. Also, we can try to solve the problem in the worst retention ratio. We can design a new campaign or extend the existed campaigns in good cohorts.
On the other hand, we can comment on the customer life cycle. For example, we can call them for the renewals at the end of the fifth month. In the customer life cycle, we can determine the right time to ask them for the third sale.
Additionally, we used some graphics to visualize the data better. We updated our graphics after the requests from stakeholders. Interpreting and finding new insights with stakeholders is the essential point of the analysis.
Originally published at https://www.linkedin.com.
