Analytics Update: Solving the Hotel Problem with Unique User Total

Localytics usage reports provide impactful insights for our customers, including session and user counts by the hour, week, month, and daypart for any specified time period. Today, we’re happy to introduce new functionality that’s making these reports even more meaningful: unique user total.

Unique user total is a new metric that can be found in the Localytics Dashboard when viewing usage reports in table view. This new metric helps customers better understand how many people visited their app during a given time period, accounting for multiple visits by the same user. For example, customers can now view the percentage of new unique users who returned to the app for another session on the same day, or the number of unique users who used the app in multiple countries in the same week.

Previously, we displayed “Total users,” which exactly summed up the columns and rows in the usage report. Now the unique user total may not match the sum of the new and returning columns and rows. This update is not a calculation error - it’s an example of “The Hotel Problem,” a common source of confusion in digital analytics. A user who visits an app twice in one day is new on her first visit and returning on her second. Previously, she was counted once as a new user and once as a returning user, creating a total user count of two. Now she’s counted as just one unique user.

New Insights with Unique User Total


The new unique user total metric enables new insights. For instance, it is now possible to calculate the percentage of new users who visit an app as a returning user on the same day. This is a great metric for gauging how successful an app is at acquiring new users. Here’s how to do it:

  1. For a given day, find the sum of new and returning user columns.

  2. Subtract the unique user total from the sum of new and returning users. The difference represents the number of users who were both new and returning on that day.

  3. Divide the previous step’s difference from the number of new users that day and multiply by 100. This number represents the percentage of new users who returned to the app at least one more time that same day.

Let’s walk through an example of this. You’ll see that on October 22nd the unique user total was 692. Adding together the new (155) and returning (569) columns for this date returns a sum of 724. By subtracting the unique users from the total users (724 - 692), we find that 32 users were both new and returning on that day. We can then calculate the percentage of new users who return to the app the same day as their first session (32/155), which in this case is ~20%.

The same rules apply when you are viewing user numbers split by other dimensions. For example, a user could open an app in both Canada and the US during the same week. Viewing the country dimension, the user would count once in the Canada column and once in the US column, creating a total of two users. Now, with the new metric, this person is counted as just one unique user (See Sep 29, 2014 row). Similarly, businesses can calculate the number of users who are in both the US and Canada during that time period. A travel app could use an insight like this to understand audience size among frequent international travellers.

User metrics are fundamental in understanding how your users engage and interact with your app. Determining “How many new and unique users came into my app this weekend?” or “How many users came into my app twice on the day they first downloaded my app?” is not a guessing game with Localytics. We’re committed to providing accurate, transparent analytics that make it easy for our customers to gain deeper insights that drive user acquisition, engagement, and retention.