Spot which clients are about to leave before they do
It finds the quiet signal, like missed sessions or slower replies, that showed up weeks before past clients left.
A sketch of what you'll get back.
Your prompt
Fill in your details. Skip any you like.
Claude and ChatGPT open with your prompt already typed in. It's copied too, just in case.
The prompt that gets copied
Attached is a history of my clients for [YOUR BUSINESS, e.g. my 1:1 coaching program], past and current, with [WHAT THE SHEET HAS, e.g. start and end dates, missed sessions]. I believe clients leave because [YOUR HUNCH, e.g. they lose momentum after month two]. First, test that belief against the data and tell me if it holds up. Then find the earliest warning sign that showed up before past clients left, ideally one I don't know about. Find it using only the older half of my clients, then check whether it would have flagged the newer half correctly, and tell me how often it was right. Finally, list current clients showing that sign now, and draft a short, warm check-in for each that doesn't sound like a sales pitch.
Upload a simple history of past and current clients. AI tests your hunch about why people leave, finds the earliest warning sign in your own data and lists who's showing it now.
Step by step
- Make a sheet of past and current clients with start and end dates, sessions, missed sessions and payments.
- Swap names for initials, then upload the sheet and paste the prompt.
- Read the warning sign it found and how often it was right.
- Send a friendly check-in to each client who shows the sign today.
What you'll get
Your hunch tested against real data, the earliest sign that a client is about to leave, how well it predicts, and a check-in note for each client showing it now.
Use initials, not full names. A short client list gives a rough signal, so treat it as a hint, not a verdict.
- Made for
- Coaches, Consultants, Small business owners
- Works in
- Claude, ChatGPT
