Reduce Churn Using Next Purchase Predictions

How do you know when to engage a customer to prevent churn?

In this post, we’ll cover the functions and use-cases for Timely Re-Engagement segments created in the Velocidi CDP. These segments are created from machine learning attributes predicting when an existing customer is most likely to make their next purchase. It uses individual purchase frequency and product category data to give a dynamic prediction for each customer.

Timing is everything

Why do we need a feature that predicts a customer’s next purchase? If you are emailing customers every day or every week without consideration for their actual purchase frequency, it’s safe to assume that most of the emails you send are ignored. If you can predict when each customer is most likely to make their next purchase, you can time your re-engagement messages to meet their needs.

Not only do you stop wasting ad spend, but you also improve the customer’s experience. The key is to catch each customer at the time when they are in-market for a product but have not yet purchased yet.

When creating timely re-engagement segments, we capture groups of customers by choosing a specific time frame in which we expect them to make their next purchase.

The standard options are:

A. Expected in more than n days
B. Expected in less than n days
C. Overdue by less than n days
D. Overdue by more than n days.

N can be any number that works best with your existing marketing strategy and sales cycles. For the purposes of this article, we’ll use 14 days in each example. But the time frames do not have to be equal.

You can also create the inverse of each by adding ‘not’ to any rule. (ex. Not expected in more than 14 days) And you can combine segments by using ‘and’ and ‘or’ rules.

There are countless ways to experiment with these rules. But we’ll go through just the most basic options as examples.

A. Your customer’s next purchase is expected more than 14 days from now – Not yet relevant to engage

The purpose of this segment is to identify a time frame in which it is not yet relevant to re-engage your customer. You can use this segment as an exclusion list in campaigns. Don’t worry about losing touch with them. All of these customers will soon fall into the “less than 14 days” segment and you’ll re-engage them at that point.

B. Your customer’s next purchase is expected in less than 14 days – Proactive engagement opportunity

The time is ripe for emails and ads! In this time frame, your customers might already be thinking about their next purchase and will be much more responsive. Remind your customers why they loved you. Make it as easy as possible for them to re-convert. Engaging customers in this time frame is a proactive move to prevent churn.

C. Your customer’s next purchase is overdue by less than 14 days – Preventative engagement opportunity

In this scenario, you expected them to make a purchase and they haven’t yet. However, the predicted date is still fairly recent. So there’s still time to win them back before you consider them churned and lost. Try sending an offer. Segments capturing customers in this time frame can be used to reactively win back customers who might be about to churn.

D. Your customer’s next purchase is overdue by more than 14 days – Probably churned

The date at which the next purchase was predicted is now long gone. The most likely conclusion to draw here is that the customer has churned. So now it’s up to you to decide if you want to consider them completely lost, or keep them on a low-frequency drip campaign, just in case.

Thanks for reading this article on next purchase predictions and reducing churn. Subscribe to our newsletter for more articles like this, plus client success stories and company news.

Related posts