Compare the cohorts correctly

I have to see how people compare cohorts. For example, if they want to understand which cohort brought in more money, they might take two cohorts — January and June, for example — and calculate the turnover value for both in July.

You can't do that! After all, the January cohort has been around for six months by July and has had time to accumulate more money than the June cohort, which has only been around for one month. 

Compare the cohorts correctly by seeing how much money each cohort earned in the same time period — in this case, one month. 

To achieve this, we use a relative month view of the cohorts, which gives us a left-aligned triangular view of the data. Then, we can compare the cohorts as rows for the same number of periods.