You're Rostering For Fear, Not For Money

Most cafe owners are rostering for a disaster that happened once, eighteen months ago, and never happened again. That single bad Saturday is still costing them fifty two thousand dollars a year, and they call it being careful.

Somewhere in the last two years a Saturday went sideways for you too. The queue backed up to the door, a regular complained on Google, and that one bad morning became the template for every roster since. Nobody sat down and worked out what the fix actually cost.

I had a client in Auckland, cafe doing about 380 covers on a Saturday, four hundred grand a year in revenue. She had five staff rostered for the 7am to 11am rush because eighteen months earlier she got burned once. Once. We ran the numbers. That fifth staff member cost her fifty two thousand dollars a year, spread across wages, ACC, KiwiSaver, the lot. Her actual bottleneck wasn't headcount, it was one coffee machine with two group heads trying to serve a queue built for three. She didn't need a person. She needed a plumber and a new machine, which cost eleven grand once and paid for itself in four months.

That's the pattern I see constantly. Owners solve a systems problem by throwing labour at it, because labour feels like the fastest lever to pull. It isn't. It's the most expensive lever in the building, and it's the one you can least afford to get wrong in a business already running on nine or ten percent margin if you're doing well, and half that if you're not.

Here’s the actual question you should be asking, and it isn’t "how many staff do I need." It’s "what’s my labour cost per cover, and where does it actually break." Pull your last twelve Saturdays. Not the one bad one you remember, all twelve. Chart covers per hour against staff on the floor. You’ll usually find the queue isn’t a people problem for more than ninety minutes of the week. Everything outside that window is you paying full wages to cover a crisis that exists for six hours a month.

I know what some of you are thinking. Understaff once and you lose a regular for good, and one lost regular over a lifetime is worth more than a season of extra wages. Fair point, and I'm not telling you to run it bare. But there's a difference between rostering for your average Saturday plus a sensible buffer, and rostering permanently for your worst Saturday because you're scared of a repeat. One is a business decision. The other is trauma wearing a spreadsheet.

The fix isn't complicated, it's just uncomfortable, because it means admitting the roster you've been running for a year was built on a feeling instead of a number. Split your Saturday into fifteen minute blocks. Chart actual covers served against staff rostered for eight consecutive weeks. You will find at least one shift, usually two, where you're paying for capacity nobody needs. Cut it. Reinvest half the saving into the actual bottleneck, whether that's equipment, a faster POS, or better prep the night before, and bank the other half. That's not a theory. That's what fixed it for the client above, and her margin moved two and a half points in a quarter without her losing a single customer.

I wrote a while back about the businesses that blame rent, wages and the economy for a death they caused themselves. Rostering out of fear is the same disease with a different symptom. You're not protecting your business by over staffing the rush. You're just paying, every single week, for a Saturday that only actually happens once or twice a year.

Do the count before you do the roster. If you can't tell me your labour cost per cover for last Saturday right now, off the top of your head, that's the problem. Not the queue.