The teardown

A gym. One site, on a retail park at the edge of a town of 60,000 people. Open twenty-four hours, 900 members, 40 a month, no minimum term. It has been trading for six years and the two owners take a combined 70,000 a year out of it.

That is enough to run all nine questions. Here they are, in order, with the numbers on the table.

The monthly picture

Membership revenue, 900 at 40              36,000
Trainer floor rent, four at 350             1,400
Vending and drinks                            900
                                          --------
Total revenue                              38,300

Rent and business rates                    11,000
Staff                                      12,000
Equipment lease and repair                  3,500
Utilities                                   2,600
Cleaning                                    1,400
Marketing                                   2,000
Insurance, software, admin                  1,500
                                          --------
Total costs                                34,000

Operating profit                            4,300

About 51,000 a year of operating profit on 460,000 of revenue. Eleven percent. The owners' 70,000 includes their own salaries, which are inside the staff line.

1. Who pays?

Members, monthly, by direct debit. 900 of them, and the payer is the user, which makes this an unusually clean business to reason about: no advertiser, no purchasing department, no third party whose interests bend the product.

Two smaller payers sit alongside. Four self-employed personal trainers pay 350 a month for the right to train clients on the floor — the gym is selling access to its members, and the trainers' clients are often members already. And the vending machine, which is 900 a month of very high margin and no work.

2. Why do they pay?

Ask a member and you will hear health. Watch the payment behaviour and something more specific is going on: they are buying the possibility of going, and the possibility is worth 40 a month whether or not it is exercised.

The average member visits about 4.4 times a month. The average member, asked at signup, intends to come twelve times. Neither number is a scandal. The gap between them is the business.

3. How much, and how often?

Forty a month, taken on the first, with no contract and no notice period. The pricing decision is more interesting than it looks.

No contract raises churn — a member can leave in January when the enthusiasm runs out. It also removes the objection that stops people signing up in the first place, and it means no debt collection, no disputes and no reputation for trapping people. The owners tried a twelve-month contract at 32 in year three and abandoned it: joiners fell by a quarter and the arguments were constant.

4. What does it cost to serve them?

Almost nothing. The marginal cost of one more member is roughly 7 a month — electricity, hot water, a share of cleaning and equipment wear. Contribution per member is therefore about 33 of the 40.

Everything else on that cost table is fixed. The rent does not care how many members there are; nor does the equipment lease, nor most of the staffing.

Strip the per-member costs out of that 34,000 — 900 members at 7 is 6,300 — and the fixed cost of opening the doors at all is about 27,700 a month. The trainers and the vending machine cover 2,300 of it before a single member joins, leaving 25,400 to be paid for out of memberships at 33 each.

5. How are customers acquired, and at what cost?

Two channels, and they are not remotely comparable.

The gym's own channel — local advertising, signage, referrals from members — spends 2,000 a month and produces about 55 joiners. That is a cost per member of roughly 36, against a contribution of 33 a month. It pays for itself in the second month.

The second channel is an aggregator: a national app selling discounted access to independent gyms. It brings in about 25 members a month, and once the discount and the commission are counted the effective cost of each is around 400.

Just over eight percent. The gym's actual churn is 4.5 percent, so the channel works — but the margin for error is a factor of under two, and January joiners churn far harder than the average. A channel that is fine at the annual average can be destroying money for the cohort that arrives in the month it matters most.

6. Why do they stay?

Four reasons, in descending order of durability. Proximity — it is on the way home, and the gym half a mile further away has lost before it opens. Habit, which is genuinely strong once formed and takes about three months to form. Sunk identity: quitting is an admission. And a small group who have forgotten, whose 40 leaves every month for a building they last entered in spring.

Churn is 4.5 percent a month. Average member life is about twenty-two months. The owners have never calculated this and would find it lower than they expect.

7. What stops a competitor copying it?

Very little, and this is the honest answer.

There is no network effect — one more member makes the gym marginally worse for everyone else, not better, which is the opposite of a network. There are no switching costs: no contract, no data, nothing to migrate. There is no cost advantage; a national chain buys equipment cheaper. The brand is a sign.

What actually protects it is the site. There are perhaps three locations in this town with the right parking, the right ceiling height and a landlord willing to take a gym. The gym occupies one of them. That is a real barrier and it is entirely a property of the lease, not of anything the owners do.

8. Can it scale?

Not in this building. The floor holds about 90 people comfortably and the peak is already near 70 on a Tuesday evening. Adding 300 members would not add 300 members' worth of revenue; it would add complaints, then departures.

Opening a second site is not scaling, it is doing it again — a second rent, a second staff rota, a second breakeven of 770 members to clear before anything is earned. The cost per member does not fall. That is the difference between a business that scales and a business that replicates, and it explains why almost every gym chain is either very large or a single site with two owners in it.

9. What would kill it?

The teardown, in one page

Who pays — 900 members, monthly, plus four trainers renting floor time. Why — access, and the possibility of going, priced against usage that runs well below intention. How much — 40 a month, no contract, deliberately. Cost to serve — about 7 a member; almost everything else is fixed, giving a breakeven around 770 members. Acquisition — 36 a member locally, 400 through the aggregator, and the second one only works because churn is under eight percent. Retention — proximity, habit and forgetting; 4.5 percent a month. Moat — the lease, and nothing else. Scale — no; a second site is a second breakeven. What kills it — a cost shock and a competitor in the same quarter.

You now have a repeatable procedure. It takes about ten minutes on a business you use weekly, and question nine is the one worth writing down.