Gyms: a cancellation is decided three months before it's sent

The billing run lands on the first of the month and there it is: fourteen cancellations. You read the names and half of them don't ring a bell, because those people hadn't set foot in the place for weeks. That's the uncomfortable truth about this business — a membership isn't lost on the day someone cancels. It was lost two or three months earlier, on a rainy Tuesday when somebody skipped a session and absolutely nothing happened. If you run a gym, a box or a sports centre, this is about catching that signal while it's still worth catching, and about which parts of the job an agent can carry without making members feel watched.

Cancellation isn't an event, it's a slope

Nobody wakes up on a Tuesday and decides to quit the gym. What happens is slower and duller: someone training three times a week drops to two after a bad week, then to one, then misses a fortnight, and one morning realises they've spent a month paying not to go. By then you haven't lost a member — you lost them six weeks ago. The cancellation is paperwork.

Which is why the day they ask to cancel is the worst possible moment to try to keep them. The decision is made, they've been quietly feeling guilty about it for weeks, and any offer you put in front of them reads as a trap. All you can usefully do that day is process it quickly and leave a decent impression.

The useful window sits much earlier — while the person still thinks of themselves as a member, still has the kit bag in the boot, still tells themselves they'll go back this week. A well-placed message then doesn't annoy anyone. It helps.

And I won't quote you the usual line about retention costing X times less than acquisition. Do your own version: last month's spend on ads, promos and commissions, divided by the number of people who actually signed up. Put that next to the cost of messaging someone already on your books.

The signal you already collect and never look at

If you have turnstiles, cards, QR entry or an access app, you're already storing the only data point that matters: how many days it's been since each member last came in. Almost nobody looks, because looking means pulling a list of four hundred people and cross-referencing it by hand every Monday, and nobody has that Monday free.

The raw number also means nothing on its own. Ten days away is alarming for someone who trained four times a week and completely normal for someone who comes to one Saturday class. Absence only means something measured against that person's own habit.

That cross-reference — personal pattern against current gap, member by member, every week — is exactly the kind of work that eats a human's morning and costs an agent nothing. Three tiers work well:

• First nudge: they've been away roughly twice their usual interval. Nothing serious yet.
• Real risk: three weeks outside their pattern. Still paying, but the gym has dropped out of their week.
• Effectively gone: six weeks or more. This message has to be different, and above all more honest.

Before switching anything on, strip out the absences that aren't drifting: anyone who told you about an injury, anyone on holiday, seasonal memberships, and recent joiners with no pattern yet. Skip that filter and your first batch sends a cheery "we've missed you" to someone with their knee in a brace. They will remember.

What to actually say at each tier

The wording matters more than the automation. Three rules: one message per tier and no more, no guilt-tripping, and always a question answerable in one line from a phone.

First nudge, light and undramatic:

Hi Marta, haven't seen you around the floor for a bit — everything OK? If your schedule's gone sideways, we've got space in the 8pm classes on Tuesdays and Thursdays.

Second tier: offer something that removes real friction. It's rarely a discount — usually it's time or company.

Hi Marta, if the routine's stopped working for you I can book you half an hour with Javi to rework it, no charge. And if now isn't the moment, we can freeze your membership for a month and you pick it up when it suits.

Freezing stings because you lose a month's fee. You lose a month's fee, not a member — the arithmetic almost always comes out ahead.

Third tier, six weeks out. The instinct is to push harder, and the right move is the opposite:

Hi Marta, it's been a while and I don't want you paying for something you're not using. If this isn't your moment, say the word and we'll pause it or cancel it, no hassle. And if you want to come back, we're here.

That reads like shooting yourself in the foot, and it isn't. People who feel trapped never come back and they tell others; people who leave feeling respected come back in January. One hygiene rule to go with it: anyone who ignores two messages in a row comes out of the sequence for a few months. A gym that pings a silent member every fortnight is spam with dumbbells.

The other leak: joiners and their first three weeks

The other big drop-off is right at the start. January and September fill you with sign-ups, and plenty evaporate before the second payment without ever building the habit. The first three weeks decide almost everything:

1. Book the first session on sign-up day. Not a welcome email — a real date with a coach, in the diary.
2. Check in on day three. Half the doubts that stop someone coming back are trivial and die in one sentence.
3. Spot the ones who haven't started. Been in once after ten days? That's when to offer the induction session, not in week six.
4. Chase failed payments the same day. Nothing emotional here, it's money: a bounced payment nobody follows up for three weeks becomes a cancellation by default, and a resentful one.

Then the obvious one: the questions about prices, timetables and class space that arrive at eleven at night on WhatsApp or Instagram. Someone asking at that hour is comparing two gyms on their phone, and they join the one that replies.

Where an agent has no business going

• Health, injuries and pain. "My knee's been hurting" goes straight to a coach. No suggested exercises, no encouragement to come in anyway.
• A cancellation someone actually means. Offer the freeze once; if the answer is no, process it without obstacles and without "pop into reception on a weekday morning". Friction there costs you a one-star review and the return visit you might have had in January.
• Anything that reads as a telling-off. "You haven't been in for 21 days", read at eight in the morning, is a jab. And don't recite exact attendance figures — attentive turns into creepy in one sentence.
• Problems messaging can't fix. If people are leaving because the floor is packed at seven, the showers have run cold for a month, or you replaced the coach who filled Thursdays, no sequence saves you. The agent will only get the complaints to you faster.

One more thing, on transparency: say at sign-up that you use attendance data to check in on people who drift. Stated up front it's a service. Discovered by surprise it's surveillance.

Start with one tier, not the whole sequence

Build all three tiers plus onboarding plus failed payments plus after-hours cover at once and you'll never know which bit worked. Try four weeks with one threshold — the three-week one, which pays best — one message, and you approving the list before anything goes out. Count how many of those people come back through the door within the next fortnight and set it against your cancellations from previous months. That's enough to decide whether to widen it or whether the problem was somewhere else.

That logic — watching patterns and firing the follow-up at the right moment — is what our operations agent is built for: it cross-references your data, drafts the nudges and waits for your sign-off before messaging anyone. You approve, it executes.