Sign in or start a free trial to avail of this feature.
9. Analyzing Customer Tenure
For a subscription business, maximizing tenure, or the length of time a customer subscribes for before leaving is critical. In this lesson, we’ll analyze the tenure length of customers of Interslice.
- Tenure refers to the number of months that a customer has subscribed for
- The tenure for a churning customer indicates the number of months that they spent as a subscriber
- In the long run, subscription companies find it easier to retain existing customers than find new ones, so targeting actions according to customers’ tenure can be an important business strategy for more established companies
Tenure Analysis for Interslice
- The average tenure for churned customers is similar to the average tenure for the whole dataset
- This suggests that there is no correlation between tenure and the likelihood of a customer churning
- This means Interslice does not need to engage in this kind of prioritizing when aiming to prevent churn
- Hypothetically, if the average tenure of churning customer was lower than the overall average tenure, it would indicate that newer customers were more likely to churn
- This would suggest newer customers should be targeted in efforts to prevent churn
In the previous lesson we analyzed trends in sign up data for customers of inner slice. In this lesson we'll analyze customer tenure or the length of time that a customer remains subscribed to the service.
In our data set, customer tenure is recorded on a monthly basis at each transaction. When a customer churns this column will tell us how many months they subscribed for.
We'll move to a report view and create a clustered column chart covering most of the canvas.
We'll add tenure month to the axis as well and customer ID to the values as well. This chart shows us the distribution of tenure for every transaction in the data set.
The distribution is skewed towards the lower values. We would expect this because we know there are a large number of new customers who have only subscribed in 2016 and 2017.
We'll add a slicer to focus only on transactions where a customer churns.
We'll position it to the right of the chart and add the MRR category field. We'll then select churn to only view transactions where a customer has cancelled their subscription. The chart now shows the distribution of the number of months a customer had been subscribed for at the time they churned.
Clearly more customers churn at lower tenure levels with fewer of the churned customers having particularly long tenures.
This broadly matches the overall distribution of customers in the data set.
Next, we'll use a new measure to calculate the average customer tenure for before they churn.
We'll select new measure from the home tab and call the measure churners average tenure.
I'll paste in the formula and run through it.
We use the calculate formula and calculate the average of the tenure month column. The filter condition is that the MRR category is equal to churn. We'll press ENTER to create the measure.
To view this average information we'll add a card to the canvas and add our new measure.
We can now see the customers churned from the company had an average tenure of 18.81 months before they churned.
We would like to compare this figure to the average tenure for the whole data set.
If there are significant differences between the tenure of churn customers and the tenure of customers more generally this can provide the company with useful information on where to focus their customer retention efforts.
We'll create another new measure and call it average tenure.
Again I'll paste in the formula.
As we can see this formula is quite similar to the previous one.
We use a calculate function with the average of tenure month column as the expression and an all function applied to the MRR category as a filter condition.
We'll press Enter create a new card and add the new measure.
We can see that the average tenure for companies in the dataset is 18.26 months.
So what does this mean.
Well if the average tenure for churners was lower than the data set as a whole, it would indicate that newer customers are particularly prone to turning and the company should focus their efforts on keeping new customers on board.
Similarly, if the average tenure of churners was significantly higher than the data set as a whole, this would indicate that large numbers of long-term customers were churning. In such a scenario the company should focus on their well established customers.
In this case both averages are quite similar This means that customer retention efforts should not focus on any particular cohort of customers but should encompass the whole customer base.
In the next lesson, we'll take this a step further and conduct a cohort analysis to look at which customers churn at specific tenure values.