A churn for the better: predicting and countering churn rate with AI
Ask a CEO how many new customers have joined her company recently, and you are likely to get a reliable estimate. Ask how many customers the company lost in the same period, and you’re unlikely to get a quick answer. However, the rate of unsubscriptions – or churn rate – is a key measure of sustainable business success. And with the right insights and strategies, organizations can convince churning customers to stay on board.
There are ample reasons why churn is an important measure to track, but the most obvious one is that it almost always costs more money to attract a new customer than to keep an existing one. No wonder telecom businesses, where churn rates can be as high as 15 to 20% annually, are virtually obsessed with churn and how to counter it. Now it’s time for businesses in other sectors to start sharing that obsession.
The reasons behind every churn
When they actively start measuring churn, many companies are surprised to learn that the rate of unsubscriptions is actually accelerating. Customers can have a myriad of personal reasons for opting out of your services, but broadly speaking, there are four categories of churn. Understanding this distinction can help your business gain a nuanced understanding of why customers are moving on, and what you can do to change their minds.
- Fake acquisitions: customers who benefited from a promotion and are now shopping for new promotions from your competitors.
- Rotational churn: customers who are hunting for promotions that apply to new customers only. To obtain the promotion, they disconnect and reconnect their accounts – sometimes through a resident family member. This phenomenon is common in telecom.
- Bad payers: customers who have had their services cancelled because of consistent failure to pay the bills. Needless to say, these are not ideal customers to target for retention.
- Real churn: loyal customers who are dissatisfied with your products or services. On average, they make up roughly 26% of all customer churn.
An 4-step approach to countering churn
Broadly speaking, successfully countering churn requires four steps: gathering insights, determining churn drivers, creating a predictive model or models, and identifying and taking impactful actions.
- Gathering insights: In this phase, the churn rate itself, the types of churn and the relevant time period within which to measure churn are determined. In addition, the goal is to collect as much info about the users/customers as possible to create a ‘DNA’ profile.
- Determining churn drivers: These insights are then pieced together in a new model to determine which factors have the biggest impact on churn, and whether their influence is positive or negative. In this step, both individual and general churn drivers are identified.
- Creating a predictive churn model: Armed with insights into these possible drivers and extensive knowledge about the clients, data scientists build a statistically sound, predictive model that determines the correlation between both. A user’s profile or specific situation will result in a lower or higher chance of churning.
- Identifying actions: These predictions allow the organization to take targeted actions to lower churn, from personalized offerings to individual recommendations and more.
4 tried-and-tested ways to counter real churn
Countering churn doesn’t always require big campaigns or investments. Sometimes, a simple call at the right time can be enough. To know which approach to take, you need to fully understand your customer and their expectations and sensitivities. Here are four basic strategies that will help you reliably reduce churn.
- Contact the customer personally: If a customer is signaling a desire to leave, contact him or her personally. Don’t be pushy, however: use the opportunity to listen to what he or she has to say, and find out if there is anything you can do to change their minds. Remember: not every churning customer is worth convincing.
- Advertise and recommend new products and services to existing customers: The best way to keep customers on board is to keep them engaged with your company. Therefore, don’t forget to keep your existing customers in the loop when launching a new product or promotions.
- Improve the user experience at every touch point: Make sure your product or service has a smooth and intuitive user interface. But don’t stop there: every interaction with your users should be effortless, from marketing e-mails to invoicing.
- Reward loyal customers: Another great strategy to keep customers on board long term is to offer unique benefits to high-ranked customers, including early access to sales or a fast track to customer services.
AI-based churn prediction for training programs at VDAB
For VDAB, Belgium’s job placement agency, delaware has developed an AI-based churn prediction model to experiment with. Based on historical data from participants and subscription data from the training itself, the model could help the agency to accurately estimate the churn rate of each training course.
“Predicting the likelihood that someone will prematurely leave the training doesn’t really help all that much. You need to know why they are giving up as well.” Says Charlotte De Vlieghere, data scientist at delaware. “To solve this problem, we added an explanatory model. Where the existing model only determines the risk of churn, the second model also identifies general churn drivers and determines how all these different factors impact the churn rate. These insights help to identify which actions are likely to have the most impact in any given situation.
“Interestingly, we discovered that only a handful of factors were enough to predict the majority of churn. In the nurse training curriculum, for example, access to daycare for participants’ children proved to be decisive. This is due to the fact that nursing training lasts 3 years and thus requires a significant commitment. In the masonry curriculum, which is a lot shorter, it quickly became clear that most students who abandon the course early did not have any prior experience. A possible solution VDAB could look into is to implement the preliminary training as an extra option in the course.”
Simon Claus, project leader at VDAB, looks back on a successful collaboration: “delaware managed to communicate the technical setup as well as the results of the predictive model in a very clear and concise way. As a result, we were able to translate these into concrete steps aimed at further improving our offering.”