Churn Diagnosis

Churn is not random. It almost always comes from one of three sources: wrong customers in (ICP miss — you sold to someone your product wasn't built for), wrong promises made (sales process miss — you convinced someone the product would solve a problem it doesn't solve as well as they expected), or wrong first experience (onboarding miss — the right customer came in but couldn't connect what they saw to what they needed, and left before they got there). Most founders treat churn as a product problem because that's the most controllable narrative. But churn is a SaaS marketing strategy problem before it's a product problem — the fastest path to reducing it is almost always stopping the wrong customers from coming in, not building more features to retain them. Paste this prompt into Claude or ChatGPT. Share churned and retained account profiles — any format works. The AI classifies churn by root cause, names the customer type to stop selling to, and prescribes exactly what to change in your ICP, qualification process, and onboarding.

1. Click "Copy prompt" above and paste it into Claude or ChatGPT to start the session.
2. Share churned account profiles: company size, industry, buyer role, deal source, time-to-churn, how far they got in onboarding, stated reason for leaving, and their use case at signup. Add 3–5 retained account profiles for contrast.
3. The AI classifies churned accounts by root cause — ICP miss, sales process miss, or onboarding miss — and names the pattern across the churned group.
4. You'll receive three specific fixes: what to change in your ICP definition, what to add to your qualification process (the one question that would have filtered them out), and what to change in onboarding.
5. The AI also identifies any churned accounts that may be recoverable with a specific intervention.
6. Use the "diagnose my churn" mode for a full analysis, or "find my ICP miss" if you want to focus on the customer type to stop selling to.

From a churn diagnosis for a B2B SaaS onboarding tool (14 churned accounts):

CHURN TYPE BREAKDOWN:
ICP miss: 9 accounts (64%) · Sales process miss: 3 accounts (21%) · Onboarding miss: 2 accounts (14%)

THE PATTERN:
Companies with fewer than 10 employees that came in through the content channel and never assigned a dedicated CS owner. They churned at 3.2x the rate of companies that had a CS team lead at signup.

THE CUSTOMER TO STOP SELLING TO:
Solo founders or 2-person operations teams managing onboarding manually. Observable at discovery: "We don't have a CS team yet, it's mostly me."

FIX 1 — ICP REFINEMENT:
Add minimum requirement: buyer must have a named CS team lead (not founder-as-CS). Remove "any team size" language from ICP doc.

FIX 2 — QUALIFICATION CHANGE:
Add to discovery: "Who on your team owns the onboarding relationship with each customer?" If they can't name a person, flag the account as ICP risk before proceeding.

FIX 3 — ONBOARDING CHANGE:
For accounts that pass qualification but have a small team: trigger a check-in call at day 5 (not day 14). The 9 ICP-miss accounts all had their last login between days 4 and 8.