Ask a room of coaches what they charge and every hand goes up. Ask what their refund rate is and the room gets very quiet.

A benchmark roundup published in September 2026 pulled together the few public numbers that exist, and they are worse than the folklore. A payments processor serving coaching puts the vertical's chargeback ratio at around 1%, one of the highest in payments. Refund requests on online courses ran 21% of sales platform-wide in Q1 2026 and about 28% in the $497 to $1,997 band, per one community platform's data, while paid challenges refund under 4%. No independent coaching-wide figure exists.

Twenty-eight percent. On the exact price point most coaches treat as their safe, scalable, "finally a real offer" tier.

The number that flips the story

The interesting figure isn't 28%. It's the 4%.

Same buyers. Often the same coach. Roughly a 7x difference in refunds depending on format.

The platform behind the data describes the low-refund format in plain mechanical terms: the paid challenge runs 5 to 21 days with a clear promised outcome, daily micro-wins, and channel-agnostic delivery so the participant chooses WhatsApp, Telegram, Discord, or email at signup. Completion lands at 70 to 80%, roughly 14 times the traditional 5% course rate.

That's the whole mechanism. Traditional course completion has been sliding for years and hit roughly 5% across most categories in 2026. Buyers who stop watching also stop believing they got value, and a refund-eligibility clock running through a 30 or 60-day window catches most of them.

The refund rate is tracking how fast a buyer does something. Not how much they paid.

Where these numbers come from, and what that's worth

Treat all of this as directional, because the sourcing is thin and the roundup says so itself. Both sources are vendors with a product to sell: a processor that wants coaching merchants, and a platform that wants course sellers to run challenges instead. The figures are the best public numbers available, and they should be read with that in mind.

The grading in that roundup is blunt: the chargeback figure is a vendor figure with no method or sample given, and every refund number is vendor data. The challenge number comes from a vendor that sells challenge tooling.

So nobody should quote 28% as gospel. But nobody has a better number either, and that absence is its own story about an industry that publishes revenue screenshots and buries refund dashboards.

The part that ends accounts

Refunds cost margin. Chargebacks cost the business.

Coaching runs a chargeback rate of around 1% of transactions, against an all-industry average of 0.26% in Sift's Q3 2025 data. Visa and Mastercard start penalising merchants around 1.5%, so a coaching business is one bad month from a monitoring program. Most coaching chargebacks are refund requests that were handled too slowly or not at all.

The arithmetic on a single dispute: the refunded amount, a dispute fee, the original processing fee which is usually not returned, staff time, and the ratio — for a $5,000 program that is easily $5,200 out and a day of somebody's time, with a low chance of winning. A refund handled in a week would have cost $5,000 and kept the account clean.

What front-loading week one actually means

Not more content. Less, delivered on a clock.

One promised outcome the buyer can name by day seven. A first action inside 24 hours of purchase, before the refund window has any emotional pull. Something visible happening daily rather than a library sitting there accumulating guilt. Other buyers in the room, so showing up is social and quitting is public.

And a refund request answered the same week, by a human, before it becomes a phone call to a bank.

The $1,997 price tag isn't the problem. The empty first week is.