What is Google Analytics Cohort Analysis?

Your Detailed Guide

Google Analytics recently added a super cool feature to analyze the delayed effect of your visitors known as cohort analysis, which is a beta version of acquisition date only. Before this new addition, webmasters and online analysts would not be able to check the delayed response of their website’s visitors. It was very difficult to determine if X visitors visited your site on Monday then how many of them visited on the next day or the day after. Google’s new cohort analysis feature will help you get and analyze this data to increase your website’s engagement.


What is “Cohort”?

Cohort is a term that’s used to describe a group of people who’ve banded together due to a same attribute. Google used the word “cohort” to define the delayed effect in analytics and create another type of time-tested segmentation to analyze user behavior. Before the feature was integrated in Google Analytics, it was quite hard to analyze the cohorts as of date acquisition, but this can now be enabled using custom variables and events.

How to Use Cohort Analysis

You can easily access the analysis feature under the audience section presented in your left sidebar in Google Analytics. Once you click, you’ll see a graph followed by a table. While the table can be quite hard to understand at first glance, don’t worry because I’ll make it easier to understand. The default graph represents the average retention rate (%) of your unique visitors over the last seven, 14, 21, or 30 days.

In the below table, you’ll see that on April 1, 2015 (third row), 174 unique users visited the website, which will be used to represent day 0. Now, look at day 1 in the third column to see how many of the 174 visitors visited the website later. On April 2, 2015, 9.2% returned and only 4.02% visited on April 3, 2015. You can check the same thing for the fourth row to find how many of 160 unique visitors visited your website on April 3rd, April 4th, April 5th, and so on.

Google Analytics Cohort Analysis

The average of seven days with a total of 1,124 visitors can be seen in the first row, which is represented in the top graph.

Google Analytics Cohort Analysis

Until now, I have seen this analysis performed on many websites. I’ve concluded that websites that aren’t performing well in search engine rankings or any other special channel for generating traffic also have very low retention rates. Websites that brand value and draw in more steady traffic boast high retention rates. It’s my hope that you can now analyze the retention rate of your website. But, the next question is where can this analysis be used? The answer is that it’s best used for analyzing websites and mobile applications.

Cohort Analysis on Mobile Applications

Due to the fact that a high percentage of the population now uses their smartphone or tablet to search the Internet, mobile applications are booming these days. That makes analyzing user behavior for mobile applications very important to continue growth. If you wonder how long users interact with your mobile app, how often users open the app in a day, or how engaging the app is, you can find all of your answers by conducting the analysis. Then, you’ll have the knowledge to make key strategy improvements that boost your company’s presence.

Likewise, whenever you make updates to your mobile application, you’ll be able to visually see the effects of the improvement. If your retention rate lowers, then it shows that you might have missed something and users don’t like the final results. You can then use your understanding on user behavior to make the next update much better. Any changes on a mobile application’s user behavior can be easily tracked and deduced to fuel your next efforts towards more engagement.

Below is an example of cohort analysis that was conducted on a mobile application with 8,908 weekly users. As you can see, the average retention rate was 32.35% on day 1, which reduces day by day. With this data, you should start to focus on how to keep users engaged with the application so that the retention rate increases with more users opening the app daily. Once it rises, there will be a higher change of getting new visitors because of mouth publicity.

Google Analytics Sessions Cohort Analysis

Configuring the Cohort Analysis Report

When you open Google Analytics to conduct your analysis, you’ll find that the report can be configured based on the cohort type, cohort size, metric, and date range.

  • Cohort Type – Currently, the beta version only allows you to access the acquisition date, thus you can see the behavior of users who visited the site on a specific date and how they behaved over a period of time.
  • Cohort Size – This refers to the change in size of cohorts by days, weeks, or even months. Configuring your report based on cohort size can help you find how many visitors visited in January and returned in the month of February. When choosing the cohort size, you can select a date range of seven, 14, 21, or 30 days while choosing the size of weeks.

Cohort Analysis Size

  • Metric – This is simply the one thing that you seek to measure. At this time, metrics can include conversions per user, page views per visitor, sessions per guest, app views per customer, user retention, goal completion, conversion, etc. All can be handy when determining the success of your retention rate.
  • Date Range – With this, you can vary the date range from days, weeks, and months depending upon your cohort size.

Cohort Analysis Date Range

It’s also possible for you to run the analysis across different segments. For instance, you could look at the average session time for visitors on a mobile device versus visitors using a desktop computer. Or, you could configure the report based on new visitor acquisitions during a certain week, such as the week before Christmas 2014. Doing this might show that your website’s visitors spend more time on the site using a desktop computer, especially before Christmas.

Summing It Up

Don’t be discouraged if cohort analysis is quite hard to understand the first time because you’ll catch on with time. It’s a very useful feature that allows you to analyze the delayed response of users directly through your Google Analytics tool. Deducing this factual data can help you make new engaging improvements to your website and/or mobile app for better conversions.

 by Shane Barker on Marketing Tech Blog