My view is that members rarely decide to leave on the day they cancel. They stopped coming weeks earlier, and the cancellation is just the admin.
So I wouldn't start with a better cancellation offer. By then it's too late, and from January 2027 you can't put obstacles in the way anyway. Start at the point where someone first goes quiet, because that's when a message from a coach still works.
The short version
- I found no reliable UK figure for class no-shows, so count your own.
- Reminders help. Text reminders raised attendance from 67.8% to 78.6% in healthcare research.
- There are five leaks: the trial no-show, the slow first month, the booked class nobody attends, the quiet member and the unfair exit.
- Watch visits, not cancellations. I'd expect visits to fall first.
- A list of people who haven't been in for two weeks is the most useful report you'll ever run.
Leak 1: The trial who doesn't turn up
Someone books a trial on Sunday night, full of good intentions. By Tuesday, the nerves have won. No reminder arrives, and nobody notices they didn't come.
There's good evidence that reminders work, though it's from healthcare. A Cochrane review in 2013 found 78.6% attendance with text reminders against 67.8% with none. And people miss things that matter: one GP appointment in twenty was lost in NHS England's 2019 figures, more than 15 million a year. A gym trial is easier to skip than a doctor's appointment, because nobody's chasing you. I'd expect the problem to be at least as big, but that's my guess.
- Send a reminder the day before, and say who'll meet them.
- Tell them where to park, what to wear and where to go.
- If they miss it, send one kind message that same day. Then one more a few days later. Then stop.
A small charge for the trial may help too. See how to get more gym members.
Leak 2: The slow first month
The gap between joining and getting a habit is where you lose people. Dr Paul Bedford's research, as reported by the software firm Gymdesk, found 15% of new members waited two weeks or more before their first visit. In one sample of 78,071, median membership was 14.5 months with at least one staff interaction, against 4.9 months with none.
That's correlation. Keen people may simply talk to staff more. I haven't read the original work, and it comes from general gyms, not small boxes. I'd still act on it, because talking to people costs nothing.
- Book the first three sessions before they leave.
- Write down one goal.
- Have a coach use their name in the first class.
- Text at the end of week one and week four.
Leak 3: The booked class nobody attends
A class with 12 spaces, 12 bookings and 9 people is a waste twice over. Three spaces sit empty, and someone on the waitlist went home.
The fix is a clear rule, stated up front.
- Set a late-cancel window, and put it in your terms. ClassPass matches studio windows of up to 12 hours, which gives you one reference point.
- Let people release their space in one tap, on the app or by text.
- Move the waitlist automatically, if your software can.
- Send a reminder two hours before busy classes. In the Nature study below, a text went out 30 minutes before a scheduled workout.
Should you charge for no-shows? On a pack, I'd take the credit. On a membership, I'd warn first, because a fee can feel like a punishment for someone who's already drifting. A member who's missed three booked classes isn't a discipline problem. They're a retention problem. Make sure any fee is in your terms, and is reasonable.
Leak 4: The quiet member
This is the biggest one. They're still paying, and you've stopped seeing them. It's tempting to leave that alone, since the Direct Debit still clears. I think that's a mistake. They'll work it out eventually, and they'll tell people they wasted their money.
Gymdesk's reporting of Bedford's work says 53% of 36,753 members visited less than once a week. Gymdesk's own benchmark puts monthly churn at 2% to 3.5% after 90 days. That's a vendor's figure, and I'd guess it isn't UK-based.
The best evidence comes from a study in Nature. Researchers tested 54 four-week programmes on 61,293 members of a US chain. 45% of the programmes raised visits, by 9% to 27%, and the top idea was a tiny reward, worth about 9 cents, for coming back after a missed workout. The lesson I take from that is simple: the moment after a miss matters. Only 8% of the programmes had a lasting effect, so this needs to be routine.
- Every Monday, list members with no visit in 14 days.
- A real coach sends a real message. Ask how they are. Don't sell.
- Offer an easy way back: "I've got a space on Thursday. Shall I put you in?"
- At 30 days, ring them.
- Offer a freeze, if they're injured or away.
Keep these texts about the person's training. Add a promotion and they become marketing, with different rules. Wording for the two-week and four-week check-ins is in our text library.
Leak 5: The bad exit
Some people will leave whatever you do. They move, they lose their job, the baby arrives. How you handle it decides if they come back, and what they tell people.
Citizens Advice says that cancelling early usually means paying the rest of the contract, but that a gym should let someone out for serious injury or illness, or a big change such as losing their job. It says that's the CMA's view of a fair contract. Fighting those cases rarely pays.
The rules are also moving. Under draft regulations due on 1 January 2027, contracts entered into from that date must let people leave as easily as they joined. They shouldn't need to contact you more than once. Feedback must be optional. I couldn't confirm the position on offers to stay, so I'd make sure any offer can't slow them down. An end-of-contract notice follows within 24 hours for online exits, or 3 working days otherwise. This is draft, and it isn't legal advice. Our pricing guide has more.
I'd treat this as useful. A polite exit question tells you what's really going wrong. Log every reason, and keep the door open.
Count last month on one page
A clipboard will do. Take your last full month and copy this table.
Worked example (made-up figures)
Here's how Tanners Yard Fitness, an invented gym with 120 members at £87 a month, might look before and after working on these leaks. The figures show the sums. They aren't results from a real customer.
| Each month | Before | After |
|---|---|---|
| Enquiries | 24 | 24 |
| Booked a trial | 15 | 19 |
| Turned up | 10 | 15 |
| Joined | 5 | 7 |
| Members who left | 5 | 4 |
| Net change in members | 0 | +3 |
Before, the gym stands still at 120. It works hard and goes nowhere. 19 of the 24 enquiries never become members.
After, it gains 3 members a month from the same 24 enquiries, with no extra advertising. After a year it has 156 members, if there's room for them. That's 36 more members, or £3,132 more a month in fees including VAT. Added up across the year, the extra fees come to £20,358 including VAT, or £16,965 before VAT.
Your numbers will differ. Look at the shape. One fewer leaver a month matters as much as one more joiner, and it's usually cheaper.
If you're losing most at the trial stage, fix reminders. If they come once and don't join, look at your first-class experience. If they join and fade, fix the first month.
Before you automate any of this
Don't make someone pay more or stay longer because you've made leaving difficult. Hiding the cancel button stopped being clever some time ago, and it's about to carry a much bigger risk.
Then think about more members
Once the leaks are smaller, every new enquiry is worth more. That's when adverts and booking apps start to make sense. See how to get more gym members for the full order of work.
Questions gym owners ask
What's a normal no-show rate for gym classes?
Nobody I could find publishes one for UK gyms. Pull a month from your booking system, split by class and time, and that becomes your baseline.
What's a good member retention rate?
Gymdesk reports a Health & Fitness Association benchmark of 66.4% kept each year, so about one in three leave. I didn't open the original, and it's unlikely to reflect a small UK box. Measure your own rate, and try to beat it.
Should I charge for late cancellations?
It's reasonable, if it's clear in your terms and the fee is fair. I'd use it for packs and PT sessions, and take a softer line with members.
Should I let members freeze?
Yes, I think so. A frozen member still has a relationship with you. Decide on the rules, such as a maximum length and whether you charge a small fee, and put them in writing.
Can I text members who've stopped coming?
Yes, in my view, if the message is about their training. Avoid offers in it. If you promote something, the marketing rules apply.
Sources
- Cochrane review CD007458, 2013 (healthcare, international)
- NHS England, 2019 (England)
- Gymdesk, reporting Dr Paul Bedford's research and its own benchmark (2 October 2026, vendor)
- Scimex, Nature megastudy (US)
- Penn Today, Wharton study (7 January 2022, US)
- ClassPass, payouts and policies (vendor)
- Citizens Advice, cancelling a gym membership
- Draft guidance on consumer subscription contracts (October 2026)
- Lewis Silkin, draft guidance (30 October 2026)
Prices and rules change. Check the source before you rely on a figure. This is general information, not legal, tax or financial advice.
