As our name suggests, we love analysing data at JimmyData. A Cohort analysis is one of those reports which is confusing at first glance but brilliantly insightful when you make the time to understand it.
A Cohort analysis used to be something that the big data guys spent a tonne of cash to have setup. Now its available for free in Google Analytics and unfortunately most marketers don’t actually know how to read one! I don’t blame them as at first glance it looks confusing. So let’s have a look at the Cohort analysis.
What is a Cohort Analysis?
To understand exactly what a cohort analysis is, you must first understand what a cohort is. In the case of Google Analytics, a cohort is a segment of users based on a specific acquisition date or date range, whether it be days, weeks or months. You can change this a bit to fit your business as well.
As an example, you may want to run a cohort analysis of people who made a transaction or a conversion on your website. To do this you need to utilise the custom segments. Once you have defined your cohort based on the date or date range and the action taken, you can run an analysis of that data.
How to Read a Cohort Report
In order to get the most out of your cohort analysis, you have to understand the report, which can be complicated. There are three main sections of the report:
- The report settings
- The tabular data,
- The data over time graph.
We will explore each of these three sections for you. First we will explain how to create a cohort report, which is the report settings. The actual report is the tabular data. The data over time graph is what you see when you are comparing cohorts or even looking at a long period of time within your analysis.
Creating a Cohort Report
When you are creating the cohort report, you must first make four big choices: the type of cohort, the size of the cohort, the date or date range, and the metric.
When choosing a cohort, you will essentially be choosing a selection from a date or date range. You will decide on the date you want within Google Analytics. This will serve as the starting date of the cohort.
Next, you will choose the size of the cohort. This is a date range. If you choose one day, the analysis will only pull data from that one specific date but as a default it starts at a daily cohort size with a 7 day date range. Depending on your product and the user life cycle you may want to have a larger cohort date range however if you product is more impulsive and encompasses a shorter buying cycle, then you can leave it at 7 days.
Naturally, if you choose a week, it will pull it for a weekly cohort size and choose from the weekly date range.
Finally you can choose a monthly Cohort which gives a much broader look at the audience. If you don’t get much volume you will probably find little use for this report however for sites with larger volumes and transactions/conversions this could be extremely useful.
As you saw in the screenshots above, the last part is choosing the date range which is predefined in Google Analytics. You can change this at any time but you will want to choose a date range for the data to pull from.
Segmenting the Cohort Report
From here you will pull the metric that you are looking for in the analysis via the custom segments. This is where people tend to get stumped in the Google Analytics cohort as instead of having a drop down to all your goals, it instead requires you to edit the custom segment to filter the data.
This is specific to your business and there is no wrong answer here as you want to filter the cohort by your key KPI. The great thing about using segments is that you are not restricted to events or goals and can instead customise the cohort to measure pretty much any action.
What this means is that you can look at data within different segments such as gender, age, mobile traffic, tablet traffic, desktop traffic, time of day, and much more. It is just another way for you to take a look at your data and analyze it in order to make better business decisions.
A good way to use this tool is to try to see if any of the segments you are looking at are performing differently than other parts of the cohort. A good place to start with further segmenting is by looking at traffic source or even campaign source. You are limited with the type of cohort you can create in the beginning but this option allows you to further define that cohort for your needs.
Where to Begin Choosing Cohorts
A big difficulty when choosing how to begin with this process is deciding which cohorts you should begin with. Again, there is no wrong answer for this question because the ultimate goal is to learn more about your business and how you can continue to make it even better and grown even more. However, if you are stuck on deciding where to start, we have a few tips that can help you explore more of what the Google Analytics Cohort Analysis can show you. Use these to get started if you are not sure where to begin and then find the other areas you want to explore.
Remember, all of the components of the Cohort Analysis from Google Analytics can be customized to meet your needs and what you are specifically looking to explore. If you have a cohort or analytic you want to explore on a regular basis, you can do this by creating a shortcut so that the report can run for you without having to include all of these steps every time. This makes it easier for you if you plan on running these on a regular basis.