Skip to main content
All CollectionsPro PlanRetention analysis
Understanding Cohort Performance
Understanding Cohort Performance
Dimira Teneva avatar
Written by Dimira Teneva
Updated over 5 years ago

Cohorts are groups of people with similar traits or behavior - for example, generations sometimes are also called cohorts.

In Marketing, cohorts are groups of customers united by a common trait - products they buy, the frequency of shopping, preferences for particular promotions, etc. This kind of categorization gives marketers much more context when monitoring their shopping behavior and makes finding behavior patterns easier and the findings - more focused and meaningful.

Now, in Metrilo, there are the following cohort analyses:

  1. By month of first order

  2. By first product ordered

  3. By coupon used for first order

  4. By campaign attributed to first order

  5. By traffic source of first order

How can you access the cohort analysis?

In the Retention Analysis tab, switch from the default Overview to Cohort (at the top of the page).

Then, the switch between the different cohort options - Months, Products, Coupon codes - is right below. 

Each one segments your customer base by the specific trait and gives you the following metrics:

  • Revenue - all-time for the whole cohort

  • Returning - the share of the cohort that is returning customers

  • Avg. order - the AOV for the cohort

  • Rev. per customer - average revenue earned from each customer in the cohort

The Cohorts by months are sorted by recency, while the other two are sorted by revenue in descending order.

Going deeper into Retention Analysis - Cohort Performance

If you click on the Performance button for any cohort, a panel expands and you'll see a matrix with numbers.

This matrix maps all next orders (after the first one) over time. It answers the question how much time after the first order these people bought for the second, third and so on time.

The columns in the matrix months/ weeks and the rows are 2nd, 3rd, 4th and so on order. 

You can look at sales in order count (#), the share of people from the cohort (%) or revenue ($). 

Lastly, color-coding will draw your attention to the periods with most sales.

How to read the matrix?

Here, we see that 29 second orders were placed in the first month and 125 more - in the second month. So this cohort didn't wait much time before coming back after their first purchase.

Then, we see most of their orders in the third month. This is a perfect example of recurring business and regular sales. It's great for a monthly subscription - and no wonder, dog food is a product people need all the time and probably buy monthly.

The fading trail you see on the 8th row forms because all orders after the 7th are calculated together and include 9th, 10th and so on. The fading is caused by churning customers until very few remain to place orders, some time after the 24-month mark. So this cohort lives to about 2 years.

Did this answer your question?