A £24 Car Wash Membership Looked Great Until Week Three
A car wash owner I'll call Dan has a simple idea. Charge £24 a month, let regulars wash as often as they like, and watch the recurring cash roll in. He prices it in his head on a drive home and launches on Monday.
By week three, a handful of members are washing every other day, his quiet-hour bays are fine, but his busiest Saturday queue is now full of people who paid nothing extra for the visit. The idea wasn't wrong. The price was a guess.
What the prompt does
The Price a Car Wash Membership and Test It Against Pay-Per-Wash prompt makes the model do the arithmetic Dan skipped. You feed it your package prices, how often regulars come, your cost per wash and your capacity. It returns three tiers, then stress-tests each one against a worst case and a best case.
A worked example
Say the inputs are: Basic wash £8, Premium £14, regulars come 1.5 times a month, variable cost £3.20 per wash, 60% of customers are regulars.
| Tier | Monthly price | Cost at 2 washes | Cost at 8 washes |
|---|---|---|---|
| Essential | £14 | £6.40 (margin £7.60) | £25.60 (loss £11.60) |
| Plus | £22 | £6.40 (margin £15.60) | £25.60 (loss £3.60) |
| Fair-use Plus | £22, capped at 4 a month | £6.40 (margin £15.60) | £12.80 (margin £9.20) |
That last row is the point. The prompt notices that unlimited washing turns a handful of heavy users into a loss, and recommends a fair-use cap before you print a single flyer. It also estimates what you lose when existing regulars switch from pay-per-wash, which is the cost most owners forget.
Where the value is
- Break-even number. You get a member count you need to hit, so the launch has a target.
- Assumptions on the table. Every guess is stated, so you can change one and re-run it in a minute.
- A counter pitch. A 40-word line your staff can say at the till.
How to use it
- Pull four numbers from your till and wash log: package prices, average visits per regular, cost per wash and hourly capacity.
- Paste the prompt into any chat model and fill the bracketed fields.
- Run the worst-case table, pick the tier with the cap, and test it with 20 customers for a month before a full launch.
Ten minutes of modelling is cheaper than a quarter of giving away washes.