The handful of numbers an escape room owner should actually track
There are two kinds of operator, and both have the same problem. One has a dashboard with thirty tiles on it and looks at none of them. The other looks at the bank balance on the first of the month and calls that management. Neither can tell you whether last quarter was good.
The useful set is small. Six numbers, each attached to a decision, reviewed once a month. Anything beyond that is a hobby.
Most of what your system reports is noise
Booking platforms report what is easy to count, which is bookings and revenue. Both move for reasons that have nothing to do with how the business is performing: a bank holiday landing differently, a corporate group that came in one line, a room offline for a week. Read them week to week and you will make confident decisions on random variation.
A useful number has three properties. It compares like with like, it is not distorted by one large booking, and it points at something you can actually change. Test any metric against those three before you put it on a wall.
The numbers that tell you about the money
Slot occupancy, split by day type
The share of the slots you offered that you actually sold, calculated separately for weekday daytime, weekday evening and weekend. A single blended occupancy figure is almost meaningless, because it averages a sold-out Saturday against a dead Tuesday and describes neither.
The decision: weekend occupancy near the ceiling means your price or your capacity is the constraint, not your marketing. Weekday occupancy on the floor means the audience is wrong rather than the price.
Average group size
Players divided by bookings, again by day type. This is the number that quietly decides your revenue under per-person pricing, and the one that moves without anybody noticing.
The decision: a falling average is a signal about your minimums, your pricing structure, or a shift in who is booking. It is also the number that makes a rising booking count a lie.
Revenue per available slot
Total revenue divided by every slot you offered, sold or not. Not revenue per booking, which flatters you every time a large group arrives and tells you nothing about the empty hours around it.
The decision: this is the single figure to judge a price change, a schedule change, or a new day-part against. If it went up, the change worked, regardless of how the booking count looked.
Lead time
The gap between when a booking is made and when it is played, tracked as a median rather than an average so one corporate booking made eight months out does not swamp it.
The decision: lead time is your early-warning system. It shortens weeks before revenue falls, which gives you time to act, and it tells you how late you can safely finalise a staff rota.
The numbers that tell you whether it lasts
Share of bookings from repeat and referred customers
What proportion of this month's bookings came from someone who had played before, or who said they were sent by someone who had. You will need to ask, on the booking form or at the desk, and the answers will be imperfect. Track the trend rather than the value.
The decision: a falling share means you are buying every customer new, which gets more expensive every year. A rising one means the rooms are doing the marketing.
Review volume and recency
How many reviews arrived this month, and how long ago the most recent one landed. Deliberately not the star rating, which barely moves and tells you nothing you can act on.
The decision: volume is a measure of your process, not your product. If it drops, something changed on the floor, usually that the ask stopped happening on busy shifts.
The traps
Three failure modes account for most of the bad decisions made from good data.
Celebrating bookings while group size falls. Bookings up eight per cent sounds like a good month. If average group size fell by more than that, you did more work for less money, and the booking count is the number that hid it. Always read the two together.
Chasing the rating instead of the count. A rating is a slow-moving average that a single month of effort cannot shift. Review volume and recency respond to what you do this week and matter more to a reader deciding tonight. Optimise the one you can move.
Reading a week as a trend. A venue with a modest number of slots has genuinely noisy weeks. One bad week is weather, a competing event, or nothing at all. Compare four-week blocks against the same block last year, and refuse to make structural decisions on anything shorter.
These numbers also behave differently once a product has no capacity ceiling. Occupancy and revenue per available slot are the right lens for rooms because the slot is the scarce thing. For a City Sleuth route, where any number of teams can be on the streets at once, the slot is not scarce and those two numbers stop meaning much, which leaves group size, lead time and referral share doing all the work. Worth knowing before you try to run one report across both.
A monthly review that takes under an hour
Same day each month, calendar entry, one page, no meeting.
- Ten minutes. Pull the six numbers for the month just ended and write them next to the same month last year.
- Ten minutes. Circle anything that moved by more than it usually does, in either direction, and write one sentence on what you think caused it.
- Twenty minutes. Read the hint logs and the debrief notes for the month, which is where the explanation for most of the movement actually is.
- Fifteen minutes. Choose one action. One. Note the number you expect it to move and by when.
- Next month. Start by checking whether last month's action moved the number you said it would.
The last step is the one everybody skips and the one that turns tracking into management. Without it you accumulate a year of numbers and a year of untested opinions.
Where to start
Do not build the full set. Pick two: revenue per available slot, split by day type, and average group size. Those two together explain more about an escape-room business than the rest combined, and both can be calculated from an export you already have.
Add the others one a month as they become useful, and stop when the page is full. If you eventually want a second product to read against the same numbers, that is what we build with operators at City Sleuth, and the rest of the operator library sits at the partner guide.