## Retention Metrics **Goal**: A set of metrics that capture retention across the entire customer journey. `What does “customer retention” really means, and why it’s traditionally a tricky metric to measure?` - Customer acquisition will almost always revolve around some sort of “sign up” or “create account” behavior. Signing up new customers is the common denominator for most acquisition metrics, and because of this, acquisition metrics are generally easier to measure. - There are many battle-tested strategies on how to conceptualise an acquisition funnel; however, **no two products will have the same retention strategy**—and this makes it more difficult to measure, since every product must define its own strategy to keep its customers. Of course, the goal of retention is the same for all businesses—don’t lose customers—but _how_ each product achieves this goal will differ. - As difficult as it is to build a compelling retention strategy, it is also difficult to measure that strategy’s effectiveness. At a high level, companies with strong customer retention tend towards profitability in the long-run, but that’s a **lagging indicator**. Profitability and growth prove that the business model is viable, but they don’t tell you _what_ worked really well. They just tell you that _something_ worked. - If we can’t use simple metrics like revenue or user growth to measure retention, then how can we do it? - As with most areas of research, it’s best to start with a simple question: **What is a retained user?** Now to answer this, you need to jump into Mixpanel, and make some observations. - Start by looking at individual customers and their activity over the platform, by looking at individual customers, its easier to eyeball retention. - Specialized tools required to create clear, specific definitions of what a retained user's activity looks like. - examine trends that are present and not present **Churn:** Opposite of retention, is often about observing things that aren't there. - Did A and came back and did B The qualifying clause and the repetition clause *Note*: You can create custom events in Mixpanel such as "meaningful activity" for the repetition clause to filter out candidates with select criteria who classify as retained customers. **Custom properties** offer flexibility to transform existing data in order to prepare it for more straightforward analysis. This metric will be far more precise, and therefore more accurate and reliable. ![[PMing_Analytics_Edit Custom Event.png]] **N day retention:** Counts for the exact days when the user came and performed a meaningful activity on the platform and also does not account for the days where user did not come and hence did not perform any activity on the platform. For eg. If a user on Day 0 performs an action lets say opens the app and orders a sandwich, that Day 0 is counted as retained. And if the same user doesn't perform any activity from Day 1 to Day 4, so those days are not counted as retained days, and if the user again comes back on the day 5 and performs a meaningful activity, that day would be counted as the retained day again. > Activity within user lifecycles = N day retention is most appropriate **Unbound retention:** This calculates the gaps as well, so as per the above example, the user will still be counted as retained for the days 1-4 as retained user and also for day 0 and day 5. - This is particularly useful because the user even though not performing any action on day 1 - 4, has still not deleted the account or completely left the platform and hence they are a retained user technically. - The only users who will not be counted for unbounded retention are the ones who never returned. - It measures how quickly or slowly we truly loose customers regardless of whether or not they come back everyday. > Churn, abandonment, or longer term retention trends = Unbounded retention is most appropriate - Ultimately retention depends on the type of metric you use to calculate it and might differ from company to company and analyses to analyses. `You need a clever and novel product retention strategy—because you’re fighting an uphill battle to keep your users engaged, and that means innovating. That means trying new things. That means running experiments.`