Why The Gym Sells More Memberships Than It Has Space

A gym with a few dozen machines can hold several thousand members without ever being unusable. The arithmetic works because most members are not there, and that is anticipated.

Capacity is planned around attendance

The number that matters operationally is how many people are in the building at once, which bears little relation to how many are paying.

Only a modest share of members attend in any given week, and the ones who do are spread across opening hours that run from early morning to late evening.

Sizing the equipment to the membership list would mean a building mostly empty, at a price nobody would pay. The model requires the gap.

Attendance is highly predictable in aggregate

Individual behaviour is erratic and the total is not. Operators know their attendance curve by hour, by day and by month with considerable precision.

That lets them set membership caps, staffing and class schedules against expected load rather than against worst-case load, in the same way airlines and hotels plan.

Access control at the door produces this data automatically, which is one reason the barrier exists even where nobody would sneak in.

Non-attendance is the profitable state

A member who visits rarely costs almost nothing to serve. They occupy no machine, need no supervision, and consume no hot water or floor space.

The same member pays the same fee as somebody attending four times a week, which means the low-usage group effectively subsidises the high-usage group.

This is not a criticism of anybody's discipline. It is a straightforward consequence of pricing by time rather than by use, and everybody in the sector understands it.

The contract is built to outlast enthusiasm

Sign-up volumes spike sharply at the start of a year, and attendance from that cohort falls away over the following months in a pattern that repeats annually.

Minimum terms, notice periods and annual commitments all extend payment beyond the point where motivation typically fades, and the length of those terms is not chosen at random.

Discounts are commonly larger for longer commitments, which is a rational trade for a member who will attend and an expensive one for a member who will not.

Peak hours are where the model shows

Between roughly six and eight in the evening the assumption breaks, and the same building that felt spacious at two in the afternoon has a queue for every machine.

Operators manage this with off-peak pricing, class bookings and equipment mixes that spread load, rather than by capping membership, which would cost revenue directly.

So the crowding is not a failure of planning. It is the visible edge of a plan that works for everybody except the people who can only come at seven.