The English dictionary defines Cohort (Noun), as:
1. An ancient Roman military unit, comprising six centuries, equal to one tenth of a legion.
2. A group of people with a shared characteristic.
We are going to focus our attention on this second definition of “Cohort” here – “a group of people with a shared characteristic”. Cohort Analysis is a fancy segmentation technique used for better understanding of user behaviour. And although it was possible to generate cohort reports in Google Analytics in the past, it did require some segmentation hacks. Not anymore.
Google Analytics finally has Cohort Analysis as part of its core feature set, in the process, joining other analytics tools like Adobe Site Catalyst which have had this feature for some time now. In today’s post, I shall discuss how to do cohort analysis and what to expect from this toolset in the future.
In simple words, Cohorts are a time-tested way of grouping people together, based on date. For example, a sample cohort could be a grouping of website visitors based on the date of their first session (also known as Acquisition Date). So for example, if a visitor were to land for the first time on a website on March 20th, 2015 then he/she will be a part of the March 20th cohort.
A visitor can also be a part of more than one cohort. So, from the previous example, the same visitor could also be a part of the “3rd week of March” cohort, or even the “March” month cohort.
Another example of a cohort might include all visitors who successfully completed a website goal within a certain time period. This is a pretty realistic scenario with e-commerce companies, where it’s common to hear the sales team talk about new customers being acquired, say, during Christmas. This is nothing but a cohort and includes all customers whose first transactions took place in the week leading up to Christmas.
Cohort Analysis entails the process of analysing these groups of people (cohorts), over a specific period of time and analysing how their behaviour might be different from other users. So, in other words, a cohort is a type of user segmentation based on time. Remember, the key here is time. Often people tend to use the term cohort to mean a segment of users while ignoring the time part of the definition. But segmenting by users, is really user segamentation. Cohort Analysis, on the other hand, does have to include time.
You might ask, what’s really the advantage in segmenting visitors in this fashion? For starters, analysing the data from such cohorts allows us observe user behaviour over a period of time and can help us to answer questions like:
1. Do these visitors really behave differently?
2. How does this behaviour differ from other visitors who buy outside the cohort time period?
3. Do they buy more than once?
4. Does the spend amount differ?
And so on. Cohort Analysis can be useful for just about any business, and not just for the ecommerce companies. A company like Moz, for example, that offers its marketing solutions with a two month trial period can use Cohort Analysis to determine how many customers who signed up for a trial membership in January went on to buy premium memberships, as compared to those who signed up in February.
If you have logged into your Google Analytics account recently, you might have noticed the Cohort Analysis (currently in beta) report. Look for it under the Audience dropdown.
Let’s take a look at the various parts of the report. The report has three main regions:
1. The settings region.
2. A graph of data over time.
3. Tabular data region.
Let’s talk about configuring a Cohort Report. From the settings region, you can choose from four different ways in which to display the cohort data:
1. Cohort Type: This lets you specify for which date you want Google Analytics to use to create the cohort report. The only option currently available here is the “Acquisition Date”. Expect to see more options in the coming days.
2. Cohort Size: Allows you to specify the time frame, thus determining the size of each cohort. The currently available options are by “day”, “week” and “month”.
So, if you choose to see “by day” option, the report would show all visitors with the same Acquisition Date. On the other hand, choosing “by week” would display results where all users had an Acquisition date within the same 7-day period.
3. Metric: This dropdown allows you to select the metric that’s being measured for each cohort. This is the actual data that you see in the report. The default value is “User Retention” which basically means users that visited more than once in the selected time frame. Other options currently available are:
• Goal Completions per User
• Pageviews per User
• Revenue per User
• Session Duration per User
• Sessions per User
• Transactions per User
• Total Goal Completions
• Total Pageviews
• Total Revenue
• Total Session Duration
• Total Sessions
• Total Transactions
• Total Users
4. Date Range: This is the time boundary that determines what data appears in the report and corresponds to the number of rows in the table below. The available time ranges are: 7 days, 14 days, 21 days and 30 days.
So, if I choose a Date Range of “7 days” and today happens to be 29th of March, then the Cohort Report will look at data from March 29th (Day 0) up to April 05th (Day 7) and create the report based on each user’s acquisition date.
Taking the above example further, this is how Google Analytics would create the various ‘Days’ of data based on a user Acquisition date of March 29th:
Day 0 = Mar 29
Day 1 = Mar 30
Day 2 = Mar 31
Day 3 = Apr 01
Day 4 = Apr 02
Day 5 = Apr 03
Day 6 = Apr 04
Day 7 = Apr 05
The following table then displays the break down of the “day data”*. Here, each data cell represents a different group of users on a different day:
*Note: We are talking of “day data” here since we chose the Cohort Size to be Day.
So far, we looked at the parts of a Cohort Report. Now let’s look at a typical workflow, taking an example. Assuming that you run a news blog, you might want to understand the behavior of your users on a weekly basis. For an information/news blog, knowing how many people are active in a certain week could give you useful insights into your content marketing efforts.
Step 1: Let’s start with selecting the Cohort Type. As mentioned above, we only have the “Acquisition Date” available as of now.
Step 2: Choose the Cohort size. In this case, if you publish new content to your blog everyday, it makes sense to use “Daily” as the Cohort Size.
Step 3: For the metric, you might want to choose “User Retention” to learn how many users return to your site every day.
Step 4: Set date range to 7 days.
And bingo! You have the Cohort report. Now for some analysis. Let’s look at the tabular data:
Since we chose a daily Cohort, each row here represents a day. The table data, therefore shows the user retention rate for each cohort for the past 7 days. Remember, each row here is a different cohort.
Looking at the numbers in the table cells, you can now start driving some conclusions. For example, you can see that users who visited the site on March 31st, re-visited the site one day later at a much higher rate (5.51%) as compared to other cohorts. On the other hand the user retention rate on subsequent days (Day1 , Day 2 and so on) seems to be higher for the cohort on Apr 1st.
Why did this happen? Was there a blog post that generated more interest? Or was it due the fact that April 1st was a holiday? Was it some campaign that was going on one of these days? The Cohort data gives you enough food for thought for your marketing activities.
Avinash kaushik once said: “Not using segmentation is a crime against humanity”. While he may have been exaggerating to some extent, there is no denying the fact that segmentation is the holy grail of all web analytics. And Cohort Reports respect this fact by allowing you to segment data. In fact, you can apply upto 4 segments to Cohort reports. Each segment creates a new table of data below the “All Sessions” table. So if I further segment this data by “Tablet and Desktop traffic”, I get:
So hopefully, this article has given you enough clarity on Cohort Reports and how you can use them to your advantage. Let me know if you have any questions.
Global Association of Risk Professionals, Inc. (GARP®) does not endorse, promote, review or warrant the accuracy of the products or services offered by EduPristine for FRM® related information, nor does it endorse any pass rates claimed by the provider. Further, GARP® is not responsible for any fees or costs paid by the user to EduPristine nor is GARP® responsible for any fees or costs of any person or entity providing any services to EduPristine Study Program. FRM®, GARP® and Global Association of Risk Professionals®, are trademarks owned by the Global Association of Risk Professionals, Inc
CFA Institute does not endorse, promote, or warrant the accuracy or quality of the products or services offered by EduPristine. CFA Institute, CFA®, Claritas® and Chartered Financial Analyst® are trademarks owned by CFA Institute.
Utmost care has been taken to ensure that there is no copyright violation or infringement in any of our content. Still, in case you feel that there is any copyright violation of any kind please send a mail to email@example.com and we will rectify it.
2015 © Edupristine. ALL Rights Reserved.