How to: User Engagement Metrics

Understanding your Power Users

Power users make the greatest companies in the world stand out from the crowd. They are in love with the product, engage with it in every possible way and bring value to the network. You want to get as many of those on board as possible. For Uber those are power drivers, for eBay – power sellers etc.. So how do you distinguish the most active users and how do you measure their engagement?

DAU/MAU method

The most common method is to determine the DAU/MAU ratio (Daily active users to Monthly Active Users) – a proportion of monthly active users who interact with your product on a single day. A common value you should consider to be good (in app world) is over 20%. If your ratio is over 50% it means either you’ve miscalculated or you’re a true genius and I take my hat off. Just kidding.

Take a look at this chart:

Customer loyalty by category

There are 3 categories that stand out. They are social games, weather forecasts and communication software. Games are used frequently at first, since you get addicted, but tend to churn quickly. The weather forecast is an interesting case since you check it only if the weather is suspicious or you plan to take a trip, so the frequency of usage is low and the retention rate is great. Communication software is an awesome example of both great use frequency and retention.

You should refer to this method if your product is both frequently used and has a high retention rate and is monetized with ads. Think about investing more time in searching for alternative methods if you want a deeper understanding of your user engagement or if your products are used episodically but the value of interaction is high (like Linkedin or Airbnb).

Alternative user engagement measurement

A couple of years back Facebook pioneered an activity histogram. It displays the engagement by the number of days users were active during the month (the length depends on the duration of a month). Generally, it reflects common activity like logins or searches but can be configured to the display any valuable metrics.

So what are the advantages of using an activity histogram over “traditional” MAU/DAU ratio?

  • Displays the user segment that returns to your product every day.
  • Shows the gradation between “just an engaged user” and a power user, unlike the DAU/MAU
  • Can be mapped to cohorts to display if the engagement gets better over time
  • Can display valuable user actions that lay deeper than traditional “app opens”.

How do activity histograms work?

It’s simple. When everything is good an activity line “smiles”. The smile can be either symmetrical or side-leaning and mean various things so it might be easier to look at an example.

Power User Engagement

The histogram above is for a news website. and it displays a clear segment of highly engaged users that return daily. Websites with high return rates like these are perfectly suited to monetize via advertisements. Products like Instagram have a “smile” leaning towards the other direction (65% return rate).

The ability to retain and grow your power users is very important. Businesses that rely on them value the network effect and expect that the users will return on a daily basis more so over time. (You can learn about the Customer Acquisition Cost here)

Let’s look at a different example:

Power User Engagement decline

This curve is totally different. It’s very heavy on the left side of and declines completely to the right. This indicates an absence of power users, which can be normal for some product categories. Companies with low engagement rates should find a way to extract as much value as possible from users the moment they are engaged, Think about a financial advisory product. There aren’t too many users that check in on a daily basis, however, businesses still make a respectable profit out of it.

Executives of these companies should think about:

  • How to create an effective revenue stream with low engagement?
  • How to get more users onboard?
  • A specific part of a customer journey (onboarding or core experience) when they lose a big portion of users because they didn’t get the value right from the start.


The timeframes

The length of an analyzed timeframe depends on your product. For example, you should analyze SAAS products in a 7-day interval. This is a valuable case if your product has a weekly cycle. Let’s take any B2B SaaS productivity application for instance (Trello, Slack etc.) in which users are engaged on weekdays and dropoff on weekends.

User engagement timeframe

DAU/MAU ratio isn’t an appropriate metric for this type of products since it isn’t intended to be used on a daily basis. By the way, if you look carefully at the curve from day 1 to day 5, you can notice a smile, which is a good indicator.

Executives of these companies should think about:

  • Who are those users that engage just once a week, but on a weekly basis.
  • What departments get more value and how can we make their workflow better so we can upsell them later?

Overlaying different cohorts can be very helpful. It can show what amount of your users become power users over time. You can notice it by a shift toward the high-frequency engagement.

Let’s look at this example:

User Engagement metrics

This curve shows 4 MAU cohorts with a positive shift in user engagement. You can notice that a growing amount of users becomes active on a daily basis thanks to a successful marketing effort that took place on November 19th and started a curve raise.

You can also measure the success of product releases and new features by looking at these curves.


What should you measure?

You can configure the activity histogram to display actions far beyond the open or login rate. For example, it can show the core activity that is related to monetization or show if the users are getting value. This is crucial because it makes you think hard about what measures are important.

This is an example of a blog platform similar to Medium. It shows the published content during the month length. You can notice a smile-like activity because while the critical mass of users publishes from time to time, there is a small group of users who do it daily. Think of yourself, how often do you post to Instagram of Facebook?

If your platform is similar, spend some time crafting the algorithm so everyone gets a chance to be read. Facebooks news feed makes sure that despite a ton of published content you get to see every post from people you care about. The same thing goes to various marketplaces – when you post a listing you want to make sure it is displayed to potential buyers.


This type of analysis allows you to get a clear view of user engagement and make more data-based decisions. You should pick a business model that is best suited for your engagement sequence. Alternatively, you can create scenarios that ignite the little-engaged audiences or invest in use cases that already brings value.

The great thing about this activity histograms is that unlike traditional DAU/MAU ratio it reflects the details of various user groups so you can distinguish what exactly drives each of them. This is the exact reason why you might want to create individual versions of these histograms per each user group.

Power User histograms display whether you have a connection with the fanatically engaged user group or no. Even if the traditional DAU/MAU ratio is low.

There is no silver bullet solution for executives to measure engagement. The only correct way is to find a sequence of metrics that work for your exact business.

Do not compare your own metrics with someone else’s even if you play in the same field. It just won’t make sense.

Thanks for reading.