Product retention is not binary

One paradigm shift initiated by the mobile app economy’s second act, which has surfaced innovative new business models that are driving massive amounts of revenue across a range of app verticals, is the idea of “churn” being temporary or even strategic. When ever more apps monetize via subscriptions, it makes complete sense that user “churn” need not be a permanent exodus from an app: a user could simply want to take a break from paying a subscription fee for some time, with every intention of later returning to the app as a subscriber.

For many developers, user retention is measured by cohort via the Day-X construction and visualized with a curve like the one above, indexed at 100% at Day 0 because 100% of a cohort is present on its acquisition day. This approach is useful because of the way Day-X retention interacts with LTV models and allows marketers to understand how DAU “stacks,” or compounds over multiple cohorts over time. The curve above is reduced by “churn,” or user departure.

Traditionally, user churn — or absence from an app — has been seen as a permanent state: if a user is churned, which is usually defined as some number of days without registering as a DAU, they are considered gone from the app for good. But as the app economy has matured and shifted into its second act, successful app developers must consider a relationship with users that might span years, if not decades. The permanent “churn” model is a relic from the time when the app economy was dominated by gaming, which is no longer the case as the Top Grossing chart has thawed: with undifferentiated games, such as hypercasual games, it makes sense that a churned user is gone for good, but that notion can’t rationally be applied universally.

The most obvious cases of this idea come from the dating and travel verticals. Churning a user is ostensibly the entire point of a dating app, but someone leaving the app temporarily is obviously not proof that they’ll never return. And booking travel is a specific desire or need at a moment in time: a person might only use a travel app a few times per year.

Some more interesting cases of strategic churn that isn’t inherent to the business model come up with streaming services. It’s perfectly reasonable to expect a user of a video streaming service to choose to churn — or stop paying a subscription — between, for example, seasons of their favorite show, or to vacillate between services as new content becomes available on each. In such a case, a user might always be a subscriber to a streaming service over the course of a year, but not one specific streaming service continuously in that time. In this case, product retention and churn are fluid concepts: not binary and certainly not permanent.

And even for some games, the Day-X retention model above stops making sense when engagement is heavily concentrated in eSports-related seasons or even weekly events, which drive a large portion of revenue for many “mid-core” games. Users cyclically “churning” out of a game only to return for events or new seasons breaks the Day-X retention model and renders it unreliable and unhelpful.

A different and perhaps constructive way of thinking about retention and churn in the cases outlined above is by segmenting users by behavior and modeling out their engagement based on probabilities of churning and probabilities of returning. This user-level probabilistic approach to modeling retention can still be aggregated up into DAU and MAU estimates, but it features the added benefit of being able to respond to cyclical factors such as new content releases, in-app events, etc.

Consider the DAU schedule above, where retention (probability of de-activation) and re-engagement (probability of reactivation) for the cohort are measured on a marginal, daily basis and not indexed against the original cohort size (as Day X retention is). This flexible approach allows for cyclical factors like in-app events, sales, new content releases, etc. to be accommodated and modeled into DAU projections.

The Day X retention model is of course very useful in working through user acquisition strategy and building broad DAU estimates, but a more granular, daily incremental view can help product teams and user acquisition teams understand the total aggregate value of cohorts, which for many apps extends beyond some short window of engagement.

Photo from Unsplash