Daily Prompt Pack: Churn Rescue – Find the exact moment customers leave and win them back

Churn usually looks sudden, but it’s rarely a surprise. The warning signs sit in your own usage data. This pack gives you three ways to find the drop-off moment and act on it before you lose the account.

[Claude]

Given your subscriber and product analytics export (list of users, cancel counts, and 2 days of usage), I need you to split the canceled users into three segments by the feature they touched last before canceling. For each pattern, write a 300-word rescue email that names the missed feature and offers one concrete re-engagement step. Output the three email drafts as markdown with a subject line each.

[ChatGPT]

Here is my churn cohort: 40 customers who canceled, with signup date, plan, last feature used, and the cancel reason they gave. Build a 3-row table showing the top cancel reasons and the earliest usage warning sign for each. Then write one 120-word win-back offer per reason, each tied to a specific metric we can fix, ending with a yes/no close. No fluff, just the table and three offers.

[Cursor]

In this repo, add a file churn_rescue.md that runs an SQL query on the events table to find users whose last-month active days dropped 60% or more versus the month before. Output that list as a markdown table sorted by MRR, then draft a three-email reactivation sequence for the top 10 accounts, each email tuned to what they stopped doing.

More daily packs in the archive: aicofounderstack.com/tag/prompt-pack