Once your data is in the right daily cohort format, it's time to conduct rolling analysis.
Since there's so much natural variance in churn data, including passive churn data (because of active churn's impact on passive churn), looking at daily, weekly, or monthly views can be misleading.
If you think about a monthly chart, what you're really looking at are cherry-picked ~30-day periods (Jan 1 - Jan 31, Feb 1 - Feb 29, etc). What this monthly view hides is the fact that Jan 15 - Feb 15 could have significantly different performance vs Jan 1 - Jan 31 or Feb 1 - Feb 29.
When viewing these cherry-picked data points, it's possible to see an upward, or downward trend, that doesn't match the actual trend line.
So once we have our daily cohorts, with In Progress dates excluded:
We sum those into rolling X-day time periods. 14- or 30-day periods is a good place to start. Here's what that looks like:
Now each row is showing a 14-day period, with summed up columns, and a recovery rate on the right which is the sum of Card Updates and Successful Retries divided by total Failed Payments during that time period.
Next Up: Visualizing Rolling Data
Once your data is formatted in daily cohorts, with the 4 outcomes broken out, and you've learned about active/passive churn, natural variance, and rolling analysis, you're ready to visualize your data.
Getting your data ready and analysis approach planned is just step 1. All of your analysis will be conducted using 3 key data visualizations, each serving a different purpose and adding unique context to your analysis.
Learn more about how to visualize passive churn performance.
