Using your in-house customer data, an AI model can alert you when a customer may be about to defect to a competitor.
Losing valuable customers to churn can deal a significant blow to your bottom line. To stay ahead of the curve, you need to get straight to the source – addressing pre-churn behaviours in your customers.
Customer loyalty is the lifeblood of businesses, driving sustainable growth, profitability, and resilience in an increasingly competitive marketplace. When customer retention costs significantly less (that is, anywhere from five to 25 times less) than customer acquisition, marketing teams should be balancing their time working on acquisition and retention campaigns. Keeping customers happy and around for longer will be much easier than constantly searching for new buyers.
This is where a retention (or customer churn) strategy comes in handy - deep customer behaviour insight through AI modelling can replace guesswork and make marketing much easier.
Churn behaviours do not exist in a “they’re doing it or they’re not” framework. At Datamine, we think of it as a scale ranging from "sliders", "soft churn" to "hard churn".
For the sake of simplicity, we will use the example of customers at a bakery chain. In this case, a high-value customer might visit the bakery daily. But suddenly, the customer doesn’t show up, or the frequency of visits starts to slow down.
What drives these behaviours, and can they be mitigated or changed using marketing?
'Sliding" denotes the slow erosion of customers over time. While they are very similar, sliders are not technically churning customers; instead, they are demonstrating pre-churn behaviours.
If the bakery’s high-value customer starts only visiting once a week (instead of daily), this is an example of sliding. The behaviour is still regular but less frequent than it used to be. Any change in circumstances could have caused the slide – moving house, changing jobs, getting up earlier to make breakfast, going to the gym, or having a baby.
The bakery hasn’t lost the customer, but this is the first step towards losing this customer to a competitor.
A soft churn customer is someone whose purchase frequency lessens over time in an irregular way. For example, high-value bakery customers stop coming daily and only visit the bakery sporadically.
Hard churn denotes a customer cutting ties with a business altogether. There are two ways somebody could hard churn:
Hard-churn customers are tough, and sometimes impossible, to get back.
Slider modelling uses statistics to detect pre-churn behaviours at a time when you can act to change customer behaviour.
Your existing customer data usually provides enough information to set up a slider model. These models give you a view of your customers’ behaviour over time, enabling you to identify patterns and detect when your high-value customers leave as early as possible. Once you have determined what data points you can use to predict sliding, you can look at individual customers to decide whether or not they look like they may slide.
The best way to stop churn is to do it before it happens, and that is fundamental to slider modelling. Using these models, you can identify recognisable and actionable critical points during a customer’s journey, from early warning signs to last-chance cut-offs.
At each stage, you can send out appropriate communications, whether through targeted marketing campaigns, offers or discounts, to capture the attention of those customers and encourage them to keep coming back.
As with any data-driven strategy, you need to have a clear understanding of the problem you are trying to solve before you can design your solution. If you’re considering slider modelling, make sure you can answer the following questions:
Most Datamine clients start slider modelling projects with their high-value customers. Typically, approximately 80% of spend comes from around 20% of your customers. Keeping your most important customers coming back for more starts by having a robust means of measuring your target population.
It’s important to ensure alignment across your teams as to how you define sliders and churn customers.
For example, X number of visits and $Y spent over a certain period disqualifies them as a "churning customer".
For example, a false positive or a false alarm could be costly for one business, while others would prefer to capture as many potential sliders as possible. A good example is shifting – for instance, utility companies often want to capture as many sliders as possible because when people move house, they typically hard churn.
Identifying pre-churn behaviours and addressing them early is crucial to retaining customers and driving sustainable growth for your business.
Create value for your sliding customers, save money on customer acquisition, and ensure customer loyalty and satisfaction by equipping yourself with a slider modelling solution. With the right approach, you can stay ahead of the curve and keep your customers coming back for more.
What’s next? Take a look at the Further Reading list below or get in touch to discuss your customer retention concerns and whether a slider model is right for your business.