How to Write a Churn Post-Mortem from CRM Export Notes
Churn memos fail when they invent that QuickBooks won and a 12 percent save rate. The CRM picklist said missing payroll. That is the reason until a human adds another field.
The matching generator is the Churn Post-Mortem from CRM Export Notes (No Invented Reasons) prompt. Browse related cards in the PromptDig library (Browse more prompts). When a filled run survives, share the version you actually use (Share a prompt).
Ledger ARR, dates, and the picklist first
Write a churn post-mortem from pasted CRM notes. No invented reasons, save amounts, or competitor names. Start by filling Inputs, not by asking the model to remember last week's run. If a field is blank, write NONE or NOT IN INPUTS and leave it blank through Generate. The card is built so the model cannot honestly invent a number, owner, URL, or command that you did not paste.
Paste these fields before you hit run:
Account and close/churn date: [Account]
Pasted CRM notes and fields (ARR, plan, reason picklist, tickets): [CRM]
People and roles in the notes: [People]
Product areas named in notes: [Product]
Save attempts I can prove: [Saves or NONE]
Competitor named in CRM (or NONE): [Competitor]
What I want decided next: [Decision]
Words I must not use: [Banned]
Word budget: [Budget or 450]
Audience inside the company: [Audience]
That inventory is the honesty ledger. Anything that does not appear there is forbidden in the draft. If you catch yourself adding a nice-to-have after the run, you are no longer using the card. You are ghostwriting. Put the extra fact in Inputs and run again.
Separate the stated reason from a resolved ticket
Generate is numbered on purpose. Do not skip a step because the first paragraph looked done. The early steps exist to stop later prose from smuggling claims.
Walk the Generate list in order:
- Honesty ledger: ARR, dates, picklist reason, tickets, people, competitor from CRM. Forbidden: reasons not in notes, save math not in Saves.
- Timeline: dated facts only. If a date is missing, write undated note rather than guessing.
- Stated reason vs evidence: quote the picklist and the free-text. If they conflict, say so. Do not pick a nicer reason.
- Save attempts: only Saves. If NONE, write no save attempt in this export.
- Product links: only Product areas named. Do not add a roadmap item.
- Decision memo: only Decision. If blank, write none requested.
- Open questions: five gaps the CRM did not answer.
- Compliance pass: quote Banned words, invented competitor, invented ARR. Cut them. Print word count vs Budget.
If a step asks for a version lock, quote the version from Inputs in the output. If a step asks for a refuse list, keep the refuse list in the published artifact, not in a sidebar you delete. Reviewers should see what the model was not allowed to do.
Record save attempts only if they exist
Most failures are the same shape: a missing field gets a confident fill. A conversion rate appears. A Gradle task appears. A flash point appears. A caption appears on a job that asked for slide text only. Your review is to search the draft for numbers, names, and commands, then grep Inputs. No match means cut.
Honor the constraints as hard stops, not vibes:
- Churn post-mortem from CRM notes, not a QBR and not a case study.
- Never invent a churn reason, competitor, or dollar amount.
- Do not write customer quotes that are not in CRM.
- Stay under Budget words.
- No emojis.
When the card says not legal advice, not certification, not an exam dump, or not a caption engine, that sentence belongs at the top of the output. Deleting it to look more finished is how you inherit risk.
Leave competitor blank when the CRM did
Finish with the compliance pass the prompt already asks for. Quote the banned-word hits. Cut them. Print character counts when the job has a cap. Print word counts when the job has a budget. List gaps as gaps. Five missing facts are more useful than one smooth paragraph.
Tags on the card (churn post mortem crm, lost account notes, no invented churn reasons) are a reminder of the job shape, not an invitation to wander into a neighboring cluster. If you need a different surface, open a different PromptDig card rather than stretching this one.
Fill the card, then run
Replace every bracket. Run on ChatGPT, Claude, or Gemini. Read the ledger first, then the artifact. If the model invents a commit, KPI, DOI, PEL, bid, or logo, discard the run. Tighten Inputs. Run again. Share the filled card that survived, not the first draft that sounded done.