How an EdTech platform decreased overall churn by 20%

In the course of a single year, more than 1.2M legitimate payment attempts failed

The company needed a solution to address the ‘leaky bucket’ of silently churning users. Silently in the sense that these users were not churning through traditional 'unsubscribe’ methods, but rather their payments were automatically failing.

For every 100 payment attempts, only 44 were successfully completed.

The company was bleeding users who wanted to pay to learn—causing pain both for the company, and for its customers who wanted to better themselves by learning new and valuable skills on the platform.

Butter’s ML-powered approach ensured recovery speed and efficiency

The EdTech company engaged Butter to handle the problem of artificially inflated failed payment rates due to improper payment configuration - timing, cadence, and presentation.

First, Butter’s proprietary algorithm sorted between legitimate failed payments and true failed payments. It was identified that ~60% of all failed payments on the platform were indeed legitimate attempts to pay — indicating willingness of an end user to use the platform, but inability to access it due to the underlying billing infrastructure.

Butter then introduced an ML-powered approach to scheduling payment collection attempts.

The company is now able to proactively prevent subscriber churn, identifying at-risk billing infrastructure and updating payment presentation accurately to ensure everyone who wants to pay to learn can.