Without any further ado, a highly engaged user is defined as someone who regularly receives value from your product.
There are 2 main elements to this definition that are worth exploring:
- the value users find in your product, and
- the frequency with which they use it
A. The value refers to the actions performed by the users (inside or outside the app) which deliver what you promised them.
What do I mean by inside or outside your app?
Well, there are 2 types of businesses:
1. The first kind are the businesses whose value lies inside the product. Take the case of Slack – an internal communication tool for companies.
The users receive value when they access the app and communicate with their teammates. The more they talk via this app, the more they are engaged with it.
2. The second type are businesses such as Zapier, which connect various apps to each other by using APIs.
It’s enough to enter this product once, make the set-up, and then the value users receive is actually outside the platform.
In this case, a highly engaged user is one who benefits of as many connections as possible out of the ones that are available on the platform.
In fact, with products such as Zapier it is very likely that users are highly engaged even if they rarely ever enter this app, due to the set-up they have.
B. The frequency is the second element of the definition. Here we are interested in how often people receive value from your product.
Naturally, the frequency differs from one product to another.
With products such as Slack, you would expect people to use it a few times a day in order to communicate with their colleagues. With the ones such as Zapier, it’s enough for users to access them once every few days or weeks or even months, according to their need to create new integrations.
To calculate if a user is highly engaged you can use this formula:
Let’s take the case of 2 users who perform 30 engagement actions in a period of 30 days. To make a fair comparison, let’s consider they are both Slack users.
The first user sends sends all the 30 messages in one day and then stops using the app. He will have an engagement score of 1 point.
This is how the engagement score is calculated:
= 1 engagement point
The second Slack user sends the same number of messages, but with a frequency of just 1 message / day. This user will have an engagement score of 30 points.
This is how the engagement score is calculated:
= 30 engagement points
Needless to say that that the second user has a much higher engagement score than the first one. He frequently uses the product and is considered highly engaged, as opposed to the first one who is probably going to churn at the end of the month.
How do you know that an action is an engagement action?
Well, when it comes to which actions are considered engagement actions and which are not, it is essential to limit yourselves to the actions that bring value to the user.
These actions are at the core of the app, they represent the very essence people are using your app for.
So, in our case, the mere fact that a user logs into Slack or someone loads the dashboard in Zapier is not enough to be considered an essential action that brings value to the user, so it certainly doesn’t make that user engaged. To perform an engagement action they either have to send messages in Slack, or to connect apps in Zapier.
How do you determine which is a high engagement score?
To find the answer to that question, all you have to do is to compare the engagement scores of various users.
So, if your second user has 30 points and most of the others have between 5 and 15, this is going to be a high engagement score, but if the others have on average higher scores than 30 points, this is going to be a low engagement score. It is that simple.
Here, at InnerTrends, we calculate all the engagement scores of the users who have churned, and a machine learning algorithm automatically creates the categories of low- , medium-, and high engagement based on the data that was collected.
User’s engagement is not static
Another aspect that should not be neglected is that engagement is not a static number, it goes up and down. A user that has a high engagement score right now could churn in a few weeks or months.
In fact, from all the data on all the companies we collaborate with, it turns out that every month they lose around 10% of highly engaged users, and their place is taken by others who come from the medium engaged users.
The engagement distribution is a quality metric for your product
Even though this is true, companies tend not to use this info when communicating with their users.
Say you’ve just launched a new feature in your app. The response from your users with respect to this new feature is different based on their degree of involvement with your product.
That new feature could secure the loyalty of your highly engaged users and prolong the time they are using your product. At the same time, it could also make a difference for the low or medium engaged ones. It might just be that marketing opportunity you were looking for to increase their level of engagement.
Amazingly enough, companies tend to communicate the launch of the new feature the same way for all the types of users, regardless of their engagement score.
For more articles on analytics, please check out my weekly newsletter. You can subscribe here.