Key Product Metrics → Deeply understood (Retention -Part 1/3)

This is part 1 of a 3 part series to talk about the product metrics that most PMs understand at a high level but not very deeply. Each part of the series will contain a detailed guide on what these metrics are, how to think about them, and more importantly how to define them.

This 3 part series will be about

• Retention metrics
• Engagement metrics
• Acquisition and activation metrics

What are Retention Metrics and why are they important?

Retention metrics are numbers that help businesses understand how many users they are able to keep over a certain period of time. This is an indicator whether users are getting value from that product. If they don’t they won’t come back.

If you have ever started your own business, you probably know how DEEPLY fulfilling it is for a customer to keep coming back to you without much (if any) incremental effort put in by you.

Why is Retention more important from the early stages itself?

One usually reads about a lot of startups flexing numbers about their acquisition — We have 1M signups in the first 6 months!! etc. It just tells whether the product is popular / has gone viral BUT it tells pretty much nothing about the business. This is where retention comes in. If users are not retained, WHAT IS THE POINT!

In fact, I would argue that retention is more important than acquisition in the early stages of a startup.

So what do retention curves look like?

In general, retention curves tell you what % of the population is continuing to use the product “x” days/weeks/months after onboarding.

There are a few types of retention curves that you might see

How do you define a retention metric?

Defining the right metric is key or else you might optimize for the wrong actions. A typical retention metric looks like this:

% users who were active in the following <time period>

There are 3 important aspects to defining the retention metric for your product:

1. Users — Define who these users are. Are these creators? Are these consumers? Are these developers etc.? There should be a separate retention metric for each user type.
2. Were they active? — A product has multiple features. Just opening up the app does not mean that it is an active user. They need to have taken a “key” step or crossed an engagement threshold bar to be considered as an active user. You need to define what this action is (We will cover more on this in Part 2 os this series). Eg- For Instagram, “hearting” an image/reel might be considered as a “key step”.
3. Time Period — This is tied to the natural frequency of the product and varies from product to product. If your product is to be used on a daily basis (such as consuming content on Instagram), then this is daily. If used weekly (such as some to-do list tool) then it is weekly. For products like Turbo tax, the natural frequency will be yearly as most people file taxes just once per year. If you don’t know the natural frequency, then make a histogram of how many days a user was active in the last 30 days and confirm whether most users are using the product as frequently as you intend them to.

Remember that retention is a lagging metric and will yield information one to two natural frequencies later.

If you are interested in building your own retention curves, you could also read this article that I have found helpful:

The Retention Curve: How best to measure user retention

In Part-2, we will talk about Engagement!