Sarah almost cancelled her gym membership four times over 18 months. Her attendance chart told the story of each near-exit — and what the gym's response (or non-response) looked like from the data's perspective.
Every gym has members who almost left. Most of the time, neither the gym nor the member consciously registers how close it got. The member got busy, the gym didn't notice, the member came back on their own. No intervention, no signal, no story to learn from.
Occasionally the data is clean enough to reconstruct it. Sarah joined a yoga and functional fitness studio in March 2023. She's still a member 18 months later. In between, her attendance chart has four distinct valleys — four moments when she came close to cancelling. Three of them were invisible to the gym. One of them triggered an intervention. The one that triggered an intervention is the one that makes the chart interesting.
Sarah's first six weeks were strong — 3 classes per week on average, consistent Monday/Wednesday/Saturday pattern. Classic new member honeymoon period. She bought a monthly unlimited membership at $155/month after her intro pack ran out. The commitment itself was a positive signal: she went from class-pack to monthly billing, which statistically correlates with longer tenure.
Attendance dropped to 1.2 classes/week across a 6-week stretch. This was mid-May through early July — a period that includes a long weekend holiday and the start of summer. The gym had no system for flagging attendance drops. No one noticed. Sarah started coming back on her own in mid-July, eventually returning to 2–3 classes/week. When asked about this period later, she said she'd been traveling for work and had genuinely intended to cancel but kept putting off the conversation. The timing of her return was coincidental — a friend mentioned a new class format and she showed up to try it.
Gym response: None. Lucky outcome.
A second dip, September through October. 0.8 classes/week — the lowest sustained attendance since she joined. This period coincided with what Sarah later described as a stressful period at work. She was paying her membership fee and not using it, which is psychologically uncomfortable — the classic "I should cancel because I'm not going" spiral. She didn't cancel. She came back to about 2 classes/week in November. The gym still had no attendance-based alerting. No outreach happened.
Gym response: None. Lucky outcome again.
The studio switched to Mako in December 2023. By early January, the at-risk member dashboard was live. The system flagged any member who had attended fewer than 2 classes in the past 3 weeks and hadn't been flagged in the previous 30 days.
Sarah appeared on the at-risk list in week 38 — late January 2024 — with 1 class in 3 weeks. The front desk manager saw her name, knew her by face, and sent a personal text: "Hey Sarah, haven't seen you in a few weeks — wanted to check in. We just added Thursday evening classes if that works better for you."
Sarah responded within two hours. She said she'd been going through a rough patch personally and had been avoiding the gym because it felt like another thing to feel guilty about. The front desk manager's response was simple: "Don't worry about it. No pressure. We're here when you're ready. Thursday 6pm is a really good group if you want to ease back in."
Sarah came on Thursday. She's been on a reasonably consistent 2–3 classes/week schedule since.
Gym response: Personalized outreach within the at-risk window. Member retained.
Another drop in attendance in June 2024 — summer again, similar to the first valley a year earlier. The system flagged her again. This time, staff looked at the flag and recognized the pattern — Sarah historically dropped in summer and recovered on her own. The decision was made not to reach out: low-intensity monitoring instead of active outreach. She recovered again in August without intervention.
Gym response: Flagged, pattern recognized, no intervention needed. Informed non-action.
The difference between Valley 2 (invisible near-exit) and Valley 3 (successful intervention) wasn't in Sarah's behavior — both dips looked similar in duration and depth. The difference was whether the system surfaced the signal and whether a human responded to it personally.
The automated email that most gym software sends to at-risk members ("We miss you! Here's 10% off your next month") does not work. Sarah was already paying for a membership she wasn't using — a discount email is noise. What worked was a person who knew her name and her schedule, reaching out with something specific rather than something promotional.
The automation's job in this sequence was narrow: surface the signal. Everything else — the tone of the text, the acknowledgment that she might be going through something, the specific class suggestion — was human. The system made the human outreach possible at scale. It didn't replace it.
Sarah has been a member for 18 months. At $155/month, she's paid $2,790 in membership fees. If she'd cancelled at Valley 2 — the most likely exit point, when she was 0.8 classes/week for six weeks and actively thinking about cancelling — the studio would have lost a member who turned out to have an 18-month LTV. They would have needed to acquire a replacement member at some cost, who may or may not have stayed as long.
The intervention at Valley 3 didn't just retain Sarah for 6 months. It retained her for the remainder of her tenure, which may run considerably longer. The text message took 3 minutes to write. The ROI math is embarrassing in how lopsided it is.
There are Sarahs in every gym's member list right now — people who are close to leaving, who would stay if someone noticed and said something personal. The question is whether your system surfaces them or whether you find out they were at risk after they've already cancelled.
See Mako in action — no sales call required
Mako is built for independent studio and service-business owners who'd rather spend their time on clients than on demo calls. Open the live demo, poke around, and see exactly how scheduling, billing, and financial intelligence come together in one place.
Try the demo: https://app.makocrm.so/demo
Self-serve. Instant access. No forms, no calendars, no "talk to sales."
A gym owner thought her admin overhead was 'a few hours a week.' The time audit showed 11.3 hours. Here's the full breakdown of where it went, which tasks were automatable, and what the gym looked like six months later when manual work had been cut by two-thirds.
A 300-member gym was paying $600/month for a bookkeeper whose primary job was telling the owner three numbers: revenue, member count, and whether cash was okay. When the gym moved to Mako, those three numbers were on the dashboard. The bookkeeper's job got smaller. Then it disappeared.
.avif)