Two or three refund requests per 100 new clients is a normal rate in the offers we run. Under five is fine. Most coaches cannot tell you their number.
They know roughly how many refunds went out last month. They do not know which closer sold those clients, which intake month they came from, or how many days after paying they asked. That is the part that tells you what to fix.
What columns does the log need?
| Column | Example | Why it is there |
|---|---|---|
| Date received | the day the request or dispute notice arrived | Speed of handling; links to the processor record |
| Client and cohort month | the month the client first paid | Shows which intake carried the problem |
| Closer | the name on the sale | Shows whether a pitch is the source |
| Days since purchase | e.g. 4, 38, 120 | Before onboarding, first 30 days, or later |
| Type | refund request, inquiry or chargeback | Disputes count against your ratio; requests do not |
| Reason in the client's words | "I was told it was done for me" | The exact words point at sales, onboarding or delivery |
| Outcome and amount | kept, refunded, dispute won or lost | Cash impact per cohort and per closer |
| Call recording checked | yes, with the timestamp | Whether the promise was really made |
A Google Sheet is enough. The closer and cohort columns come from the sale, so pull them from the client row rather than typing them twice.
How do you count it?
Per 100 new clients, never as a raw count. Six refunds in a month of 300 new clients is two per 100. Six in a month of 60 is ten per 100. The raw number looks the same and means the opposite.
- By closer: requests and disputes per 100 clients each closer sold.
- By cohort: requests and disputes per 100 clients who first paid that month.
- By timing: before onboarding, inside the first 30 days, after 30 days.
- By type: refund requests and disputes in separate columns, since only disputes hit the processor ratio.
What does each pattern mean?
| Pattern in the log | Points at | First fix |
|---|---|---|
| Spread evenly across every closer | The program or onboarding | Onboarding speed and the first win inside two weeks |
| Piled under one closer | That closer's pitch | Pull the recordings, play them to the closer next to the agreement |
| Many requests before onboarding | Fear right after the sale, or an oversold call | Onboarding call booked within days of payment |
| One cohort much higher than the rest | Something that changed that month | Check launch messaging, a new closer, an offer change |
| Disputes with no refund request first | Clients could not reach a person | One named contact for refunds, answered within a day |
The last row matters most for the processor. A client who asked and got a call rarely goes to the bank. A client who never reached anyone often does.
I track it over time as well as live. If you only see the current month, you never see whether it got better or worse three months ago. Then you loop in the same place and never improve dramatically.
Who keeps the log, and who reads it?
One person takes every refund request, and that person keeps the log. In the offers where we put one person on refunds, refunds went to about one a month or less. The same person sees every reason, which is how patterns get noticed at all.
The founder reads it once a month next to new clients by closer. The sales lead gets the closer view. Client success gets the timing view, because requests inside 30 days are an onboarding problem as often as a sales one.
Watch out. Do not use the log to punish a closer on one bad month. Ten clients is too few to judge. Look at three months, then listen to the calls before you say anything.
Diego Miescher
