How often are users using this feature per day?
Last time we looked at a product feature from a habit perspective. We wanted to figure out whether there is some percentage of users who use our product feature for a whole week (5+ out of 7 days).
Today we will look at a product feature from a daily intensity usage perspective.
To get a sense of the product feature our users should play with it. For some users to get some of the value of a product feature is just enough to use it 1–2 times per day. For other users (who probably have specific needs) to start using a product feature quite intensively from the beginning is normal behavior.
This signal could give us a few interesting pieces of information:
- we can learn what’s casual usage of a product feature per day.
- we can learn whether there is some percentage of users who get extreme value from using it (or using it in a nonstandard way).
Let’s start by recalling our case: we decided to improve chat visibility, so we changed the app home screen layout and made the chat button more visible.
As wrote in one of the previous posts [# events] = [# users] * [# events per user].
Let’s check the overall dynamic of metrics [# events], and [# users] for a chat product feature.
As we can see on the chart above the metric [# events] is growing but not as fast as the metric [# users]. It means that the metric [# events per user] should start slightly decreasing.
Let’s check the metric [# events per user].
It looks like it’s indeed declining a little bit, but let’s not jump to conclusions.
The first rule of thumb is to slice the metric [# events per user] for new and old users separately.
It’s really important as old users, as a rule, have already had some experience with the product (or some particular features) and that’s why they most like will use it in a more intensive way.
On the left side of the chart below is_new_users = False, that’s our old users. On the right side of the chart below is_new_users = True, that’s our new users.
As we can see from the chart above the intensity of usage for new and old users is quite different:
- median of daily intensity usage is much more close to the behavior of new users. It means that the new user group is dominant and shapes the trend.
- At the peak, new users use our chat feature ~3.7 times per day.
- At the peak, old users use our chat feature ~5.6 times per day.
By itself [# events per user] ~ 2.8 for chat feature for new users is a good result. It could mean that users quickly get the value of it.
The second rule of thumb is to build a histogram.
Let’s build a histogram intensity usage for new users.
From the chart above we can see that:
- Only 58.5% of new users used a chat feature just one time per day
- Some new users used the chat feature up to 50(!) times per day
- ~83.4% of users use the chat feature up to 3 times per day, which means that 16.6% of users use the chat feature more than 3 times per day.
Now it’s time to calculate histogram intensity usage for old users.
From the chart above we can see that:
- Only 36.9% of old users used a chat feature just one time per day
- Some old users used the chat feature up to 50(!) times per day
- ~62.2% of users use the chat feature up to 3 times per day, which means that 37.8% use the chat feature more than 3 times per day.
SUMMARY:
- Daily intensity usage is a signal about how good a specific feature is for users.
- Daily intensity usage should be analyzed separately for new and old users.
- The best way to get insights about daily intensity usage is to build a histogram.
Having this information at hand we can track whether making changes to a product feature will improve daily intensity usage or not. And tracking the percentage of users who use our feature intensively we can see whether our number of core users is increasing or not.
