Determining Our Usage Interval

I want to know how often I can expect people to use my app. I want this understanding to be derived from actual data, as opposed to my best guess of what I think is ‘normal’.

Once I’ve established a baseline for how often the majority of people have meaningful interactions with the app, then I have a clear benchmark to help people get to with my onboarding efforts.

I intend to plot my usage interval in Amplitude since it’s a tool designed specifically for this.

This means:

  1. Signing up to segment and Amplitude. Adding a tracking snipper to your app so that you can capture events. A connecting segment to Amplitude. They have a nifty little debugging tab so that you can make sure events are getting to Amplitude.
  2. Tracking incoming users. This means figuring where in your authentication flow you can send an event with basic user data to establish who you are tracking.
  3. Tracking your critical event. This means figuring out what your critical event is. Your critical event is the thing you want people to do on your app. The main thing that helps then derive value from your product. For client tree this was helping people, so marking a helpful task complete.
  4. Create a usage interval chart in Amplitude. It’s one of the 10 charts they offer, once you are tracking users and a critical event then its easy t set up.

On day one my usage interval chart looked like this:

preview-full-Screenshot 2019-10-15 at 12.09.22 PM.png

This is what it should look like once we get some users using the platform…

preview-full-Screenshot 2019-10-15 at 5.16.08 PM.png

So I’m measuring how long it takes people to help someone after they help someone the first time. Basically it tells us how frequently our users help people. the interval between the first and second helpful task tell us the usage interval. So in the example chart above most people (62%) do the second task 7 days out. So then we know that we have to strive to get people to use the app once a week to keep them active.

I managed to get two other charts up to help me track retention

preview-full-Screenshot 2019-10-15 at 10.10.43 AM.png

It only tracks one event per day per user, so if you come back in a month it should look more like this…

preview-full-Screenshot 2019-10-15 at 5.13.06 PM.png

This chart is important because it tells us what percentage of people who sign up stick around and come back to use the app more than once.

This is why the previous app I worked on died a quick death. So tracking retention is my primary concern now.

The last chart I got up is a retention lifecycle chart.

preview-full-Screenshot 2019-10-15 at 12.06.34 PM.png

Not very impressive right now but when we start getting users it should look a little more like this…

preview-full-Screenshot 2019-10-15 at 5.21.40 PM.png

This shows the proportion of new users to current users to dormant users to resurrected users. This way I will know if we need to focus on reviving dormant users or onboarding new users, and more importantly, I will see what effect changes we make have on these numbers


Now read this

Optimizing Retention

I got accepted to CXL institutes’ conversion optimisation mini-degree scholarship. It claims to be one of the most thorough conversion rate optimisation training programs in the world. The program runs online and covers 74 hours and 37h... Continue →