How to Improve Customer Retention with Churn Prediction

Customer retention is an important part of long-term success for any business and especially so for SaaS.

Monthly fees and other payments provide a stable source of consistent revenue to keep your business on track.

In fact, a study by Forbes found that 80% of a business’s future revenue will come from 20% of its existing customers. Since these customers account for such a significant portion of revenue, keeping them coming back is essential for future success.

Customer retention is also much more cost effective than new customer recruitment. A study found that recruiting new customers can cost five times more than satisfying and retaining existing ones.

The importance of customer retention for SaaS is clear, but how do you go about reducing churn and retaining valued customers?

Focus On Customers at Risk of Defecting to Maximize Retention Efforts

One strategy to keep existing customers is churn prediction. Churn prediction is a method that analyzes past customer data to identify and target customers who are likely to stop subscribing to your SaaS. Churn prediction is a valuable customer retention strategy because it focuses on customers at risk of defecting and helps to maximize the effect of your retention efforts.

Here are the three key steps of churn prediction.

#1 – Collect Data to Understand How Each Factor Effects Existing Customers

The more information you have about past customer behavior, the better you will be able to predict the actions of current customers.

Even information you might consider irrelevant can help inform your predictions. SaaS data such as time spent on service, number of times logged in and last time since last log in can be used to predict future customer behavior.

Other information such as number of interactions with customer support and satisfaction after these interactions can also be valuable for predictions.

Collect as much past data about as many customers as you can.

Organize this data into a chart sorted by customer and by number of months in the past. Data about the past is important for understanding how each factor affected customer actions at a later point.

#2 – Upload Data to a Prediction Service to Reveal Trends

Now that you’ve gone through the effort of collecting and organizing data about your past customer behavior, it’s time to generate a model that tracks behavior patterns.

Prediction services like BigML or Google Prediction API will create a model to reveal trends in past behavior and the resulting churn rates. These online services make it easy for even small organization to use churn prediction. Simply upload your data file and run the program to create a prediction model.

#3 – Start Making Churn Predictions to Retain Your Customers!  

After generating your model, add current data about customers to predict future churn. By comparing current customer data to data about past lost or retained customers, the prediction services will be able to identify current customers you are at risk of losing in the future, while also identifying how far in the future they are expected to defect. These customers can then be targeted to prevent defection from your SaaS.

Keep Your Customers

Maintaining subscribers is crucial for the success of your SaaS. By using churn prediction, you can identify customers who are at risk of defecting and take the appropriate action to turn them into loyal, long-term customers who will keep your business thriving.



Sydney Rootman is the Editor and Lead Writer for SaaSAddict. SaaSAddict shares news and information on SaaS, cloud migration and product marketing, in hopes of fostering discussion and interaction with the professional community.