Member intelligence for a boutique fitness network

The business

A boutique fitness brand operating across multiple locations, with several thousand active members paying monthly subscriptions. Each location runs its own mix of classes, personal training, and open gym access.

The business was growing — new sites opening, membership climbing — but the operational infrastructure hadn't scaled with it. Each location had its own booking system, its own spreadsheets, its own way of tracking who was showing up and who wasn't. Head office had visibility into revenue and not much else.

The challenge

The problems weren't dramatic. They were slow, quiet, and expensive.

Churn was invisible until it happened. A member would stop coming for three weeks, then cancel. By the time anyone noticed the pattern, the member was already gone. There was no early warning system, no risk scoring, and no way to intervene before the cancellation email arrived.

No-shows were a known problem with unknown cost. Classes were filling on paper but running at 70% actual attendance. Members who wanted spots couldn't book them. Members who had booked weren't turning up. Nobody could quantify the revenue impact or identify repeat offenders.

Location performance was opaque. Some sites were clearly busier than others, but nobody could explain why in data terms. Was it demographics? Class mix? Pricing? Management was making decisions based on anecdote and site visits rather than numbers.

Staffing was based on schedule, not demand. The same number of staff worked Tuesday mornings as Thursday evenings, regardless of actual foot traffic. Payroll was the biggest cost line, and it wasn't aligned to when members actually showed up.

What we built

A command centre that consolidates member, booking, and financial data from every location into a single operational view.

Network-wide dashboard. Total active members, revenue, churn rate, and visit frequency across the entire network. Revenue breakdown by location with month-over-month comparison. Alerts surface anomalies — a location where membership dipped 5% in a week, or a sudden spike in late cancellations.

Member analytics. Age distribution, gender split, average tenure, visit frequency, and peak hours across the network. Every metric can be filtered by location to spot differences. The peak-hours analysis alone reshaped staffing at three locations within the first month.

Individual member profiles. Every member gets an engagement score based on visit frequency, recency, and booking behaviour. Churn risk is flagged based on declining patterns. Each profile shows visit history, identified risk factors, and recommended retention actions — so staff can pull up a member's profile before a conversation and know exactly where they stand.

Class utilisation heatmaps. A weekly grid shows every time slot at every location, colour-coded by capacity. Instantly visible which slots are overbooked, which are underused, and where scheduling changes would have the most impact.

Acquisition and leads. Marketing spend by channel, lead pipeline by stage, signup source attribution, and cost-per-acquisition by location. Connects the top of the funnel to actual membership outcomes.

Financial overview. Revenue, costs, and margin by location in one view — the first time the business had a consolidated network P&L without manual spreadsheet assembly.

What the platform does

Churn risk surfaces early. Engagement scores and declining-visit patterns flag at-risk members weeks before a typical cancellation window, so staff have context in hand before the retention conversation.

Class scheduling has an evidence base. The weekly heatmap makes it obvious which slots are overbooked and which are running empty, turning timetable decisions into a data question rather than a judgement call.

Staffing can be aligned to actual demand. Peak-hour patterns become visible at location level, so rota decisions stop being based on a fixed template and start reflecting when members actually show up.

Locations become directly comparable. Any two sites can be put side by side on the same metrics, making it possible to see what's driving performance differences and replicate what's working.

Concept build. Fictional brand. All data illustrative.