Mark hasn’t been in since March. His card still runs on the first.

Mark joined in January. Came in four times, got busy, hasn’t been back since March. His card still gets charged on the first of every month, and if you pull up his location’s numbers you’ll see a healthy member count and steady billing. Mark is still a member. He just forgot.

Every membership business has some Marks. Car washes, fitness studios, massage, you name it. I don’t know what the share is at your brand, and I’d be suspicious of anybody who quotes you one without looking at your data. But it isn’t zero.

I owned two massage locations, so I know what a steady first-of-the-month draft feels like. That money pays the rent, and you don’t go looking for reasons to question it.

Franchising adds a wrinkle. Royalties usually run on that revenue, so the brand gets paid on Mark too. Nobody’s report gets upset. A unit with thinning visits and steady billing looks an awful lot like a healthy unit on the royalty report, for a while.

So why bring it up now? There’s a provocation making the rounds this fall that a lot of revenue quietly depends on human error, subscriptions nobody remembers to cancel being the obvious one, and that AI agents working for consumers will clean a lot of it up. “Go through my card statement and cancel what I’m not using.” It’s a pretty easy thing to imagine asking for.

I don’t know how fast that shows up. I’d want to see it in real numbers before I told anybody to panic. And some of it will depend on how your cancellation path works, because an agent can’t do much with a process that requires a phone call during business hours.

Somebody is going to read that last sentence and think the answer is more friction. I’d stay away from that. I’m not a lawyer, and cancellation rules are a question for your counsel, but even setting that aside, you’d be making a long-term bet against your own customers to protect people who’ve stopped showing up.

What I’d expect, whenever it comes, is that it won’t arrive evenly. A churn number moves a little every month, and everybody has a story for it. A unit that’s been living on Marks could lose a bunch of them in a hurry. And that’s a much worse day to find out than today.

How well you can see this right now depends on what you run. Some of you are in good shape. A car wash with plate reads can know how often each member comes through. A studio brand on one shared booking system can pull visits by member across the network. If that’s you, the data exists, and the question is whether anybody puts it next to revenue, by location.

Others have visits sitting in each owner’s point of sale, and headquarters sees member count and churn and not much else. And some models don’t map cleanly at all. On a home service plan nobody is supposed to show up every week, so what you’d watch is whether the visits you owe those members are actually getting scheduled.

Here’s my thought starter, and your model will change the details. Put one trend next to revenue for every unit: visits per paying member. If you want a sharper version, count the paying members with no visit in the last 60 days, and adjust the window to however often a good member actually comes in. Flat billing with a sliding visit trend is the unit you want to call, while it’s still a coaching conversation.

Now, owners are going to push back on this, and they’ve got a point. You call Mark, Mark remembers to cancel. Some of them will. I don’t have a clean answer for that. The best I’ve got is that a member who comes back is worth more than one who’s still paying and thinking about you once a year when he looks at his statement, and that if the agents do show up, you’d rather have had that conversation yourself.

It also matters who sees the number first. If this lands as one more way for the brand to grade a unit, you’ll get a lot of arguments about the definition of a visit. Give it to the owner first, as something that tells them about their own business. I spent years building analytics for health and wellness businesses, and the stuff owners actually asked for was the stuff that told them something they couldn’t get anywhere else.

If you want to try this, pick five locations and ask for the members billed last month who haven’t been in for 60 days. If you get it by Friday, put it next to revenue and see which unit surprises you. If it takes three weeks and two vendors, you’ve learned something too.