Can you see churn coming?

Predict (and prevent) customer turnover

What if you could tell customers were about to leave before they knew themselves?  This knowledge could be one of the most valuable tools in your retention toolbox.

Predicting and preventing churn involves pulling together customer data and other information from across your business, building predictive models and using that information to create prevention strategies.  By profiling the customers at risk of leaving, you get valuable insight into the causes of churn – and valuable time to reach out before they go.

Does it all sound a bit complicated?  That’s where the Datamine team comes in.  As data collation and modelling experts, we know how to handle your data and get the most value from your results.

Here’s how it works:

Churn prediction modelling

You don’t predict customer churn by reading tea leaves or going with your gut – it’s all about data.  Predictive modelling starts with a massive collation of all your business data: from your CRM data and digital marketing metrics to sales information and email list data.  The more data, the better. At this point, variable inputs are added to ensure modelling reflects the reality of your business. All that information is used to build predictive models, linking customer behaviour and other factors to churn.  Essentially, you’re creating a list or breakdown of signs that show churn is likely.

The next step involves testing and validating these predictive models using historic customer data.  By inputting information from customers with known outcomes – that is, those who have already left – you can see whether your churn prediction model is throwing up false positives or missing key churn indicators.  Of course, some misses are going to happen – the key is to understand the rate of missed opportunities and know how much it’s likely to cost your business.

Customer insights, tailored messaging

Your churn prediction models have been built, tested and validated – now, you have a valuable opportunity to stop churn before it happens.  When you know which customers are likely to leave soon, you can develop ways to reach out and keep them around.

Great modelling helps give you insight not just around the likelihood of churn for a particular customer, but also into the customer segments you’re targeting and even their reasons for leaving.  This makes the creative process easier to finesse.  Whether you create an eDM journey, an SMS outreach, customised digital ads or a combination of different elements, you’ll be able to tailor messaging to suit the segment.

Run and measure

The next step in your churn prevention journey?  Running your outreach initiatives and testing their performance.  This key part of the process helps you refine your churn responses and work out the best ways to target at-risk customers.  Each initiative needs to run for a significant period, so you get real information about its value.

Measurement involves two different elements: whether your modelling is effective and whether specific outreach efforts are working.  Essentially, you track retention in four different groups: those likely to churn, those without churn indicators, those who have received communication, and those who have not. 

The results help you assess the accuracy of your modelling: if the group at risk of churn has a lower retention rate, it’s on track.  They also help you work out if your prevention work is worthwhile – how many customers are sticking around thanks to your outreach efforts?  Are different channels or messages more effective?  With this information, you can decide whether churn prevention is worth the time and money investment – or move on to other, more effective strategies.

Predict, prevent, retain

Everyone wants to keep customers around – after all, they’re the lifeblood of your business.  While some churn is inevitable, it is possible to reduce customer loss.  Churn prediction, followed by strategic prevention and measurement, is one method.  Done well – with help from data experts – it can be an effective way to gain insight into customer behaviour, customise communications and keep them coming back.

Talk modelling and churn prevention with the Datamine team.