DAU/MAU (Sticky users) Measurements in Google Analytics

Tagged as: products web cloud

Measure the stickiness of the users of your product. Do you need to know how many of the same users use your product Daily ? The usual Active Users report doesn't give this numbers. It just tells how many users used the product in a given period. Say, you take two consecutive weeks, do you want to know how many of the users are same in this period. Such users are called WAU (Weekly Active users), they come to you atleast once every week. Similarly goes DAU (Daily) and MAU (Monthly).

If we want to do this, it requires some extra configuration in Google Analytics Reporting. You also need to send some unique ID for every user from your product to Google Analytics with every event/page-view.

Be aware that Google Analytics Terms clearly disallows we sending personally identifiable information (PII) to it, so don't use email, twitter-handle etc as this userId. Typically you need to register this user with your servers and send an internal ID from your server as the unique User Id. (You might at this point wonder, then why don't I collect analytics on my server itself - too much server load and absence of reporting or calculation mechanisms are the reasons here)

So the steps:

  1. Setup (5 min work) - Reference section below has detailed steps
    1. Add a custom dimension in Google Analytics (let us call it some name, say dimension1). Admin will have to setup the dimension.
    2. Then change the tracking code to send the userId as this dimension.
  2. Give it one or two days, so that you have enough data to analyze.
  3. You can now see this dimension1 in many places in the report. Custom reports, Secondary dimension in other reports etc. More details in the references.

If you do till here it self, you get lot of value. You can do things like get a list of users (userIds) who visited your purchase page, but couldnt complete the transaction. You can download the CSV from Analytics report and might offer a help email to them. There is no privacy compromise here as you are not sending emails to external system like Google Analytics, but just the Ids. Once you get these Ids, you can use your server's database to map the emails (which you anyway had).

Calculate DAU/MAU/WAU:

  1. Query the report by two dimensions: dimension1, date and some metric (say, sessionDuration). Give a date range of say 6 months. You can do this using the Google Analytics Query Explorer and export as TSV from the tool. If you are a developer use the API directly. You wont get this in the Analytics Dashboard. Query explorer is linked below.
  2. The above result will give you one row for each user's day. So a count of this is the number of days he used the product in the 6 months (180 days). Say the user had 24 rows, then he used for 24 days in the last 180 days. This 180/24 is approximately 7 (one in 7 days) and he is a Weekly Active user (on an average). Similarly we can compute MAU. These calculations can be done in spreadsheet pivot tables (or programmatic for a developer)
  3. Key inputs to the Query Explorer tool are : dimensions, metric, sort (important when there are lot of data) and dates. See one example Query Explorer configurationn

References: Google Analytics Custom Dimensions Setup, Custom Dimensions Reports, Google Analytics Query Explorer

Tagged as: products web cloud
comments powered by Disqus