Measuring premium churn at a global subscription business (Part 1)

Otto Szoke
Aug 22, 2017 · 5 min read

Churn might seem like one of the hardest KPIs to influence (especially for a subscription business!). Here at busuu, analysis has helped us to understand the levers and act upon them to solve customer problems and reduce Premium churn (our paying customers cancelling, both actively and passively).

We had already committed ourselves to a data-driven approach to user acquisition and optimizing the funnel, but we had not yet applied the same principles to churn.

Firstly, churn is strongly related to the quality of product — there is no doubt about that. If your product is weak and the platform crashes every 2 minutes, you are bound to have high churn.

We split the deep dive into two sections:

  1. Gain understanding
  2. Find and implement quick wins

We started by thinking about our churn metrics and what the main drivers are. Here at busuu, we offer 5 different subscription plans and track successful renewals for each. We focus on 1-month subscriptions as one of our most important metrics and we monitor this on a daily basis. The reason we chose 1-month subscriptions is so we don’t have to wait 12 months to see if our customers with an annual subscription recur or not. It looks something like this:

Insightful graph, huh? 😒 The usual commentary on this was something along the lines of “Recurring success dipped”, or maybe “It improved in week 5”, with no one really knowing what had caused the fluctuations and whether we should be concerned.

So we started looking at drivers of these dips and increases. Firstly, the supply:

The fact that you can buy a busuu subscription around the globe does make some analysis quite hard, but the key messages are…

If we have a good week in countries 1, 5 & 9, our recurring rates are likely to be lower that week. This is just how our business operates. This was particularly helpful for general awareness and there were no longer faces likes this 😧 or like this 😬 when faced with a sudden dip in recurring rates.

And yes, we are also working on improving churn in the low-performing regions. However, from our previous analyses, we are aware that in some countries, customers just want to try busuu for a month, and that’s okay too!

Then we started to look into trends within each country and it looked something like this:

At a first glance, it looks like a mess, but it’s actually quite insightful. The observant people reading might have noticed that the gap between the best and worst performers has narrowed significantly. Secondly, from the data validation perspective, the findings match the previous graph 😼 (the same countries being the worst performers).

We also looked at payment methods 💳 to address passive churn (i.e. failed recurring payments) but we couldn’t see any correlation there, so I won’t bore you with another graph…

We were also curious as to whether users purchasing a discounted busuu Premium subscription during a promotion period are more likely to churn. The widespread assumption in the business was that these users are most likely just trying the service and then cancelling straight away… but our data showed the complete opposite.

So, what about the marketing channel the customer came through? Does that make any difference?

Quite surprisingly our PPC channel performs best — well done busuu acquisition team 💪 for getting good quality customers in! This is all very well and good, but what can we actually do to reduce Premium churn?

We had many discussions around sending “win-back” and “nurture” emails before users’ cancellation dates, but further data drilling actually showed this interesting fact (my favourite one from the deep dive 😎):

Yep, that’s right! A lot of our Premium users cancel within the first 5 days of their 1-month subscription. By the way, we also ran this deep dive for other subscription plans (12-month and 24-month) and it showed exactly the same story.

In my next blog post, I will go through a few examples on how we solved this problem using customer feedback and how we improved product processes.

We have plenty of ideas already on how to reduce churn even further, such as more improvements to our web cancellation flow, as well as proactively rewarding our most loyal customers. Churn is not something you want to leave unwatched, so I’m sure you’ll want to do something similar in your business.

What does churn look like in your company? Are you prioritising it as one of your main KPIs? What works and what doesn’t? We’d be happy to share ideas.

)

Otto Szoke

Written by

Business Intelligence at busuu

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade