Most cafe menus aren't built on data. They're built on what the owner likes making, what looked good on someone else's board, or what's always been there. That gap between what feels right and what the numbers actually say is where margin quietly leaks out, month after month, without a single visible change to the menu.

The number that should worry you

Most Australian cafes net somewhere between 3% and 6%. The operators clearing 8% to 10% aren't doing anything exotic. They're not in a better location and they're not selling a fundamentally different product. What they usually have is a stronger retail or wholesale coffee component and tighter roster management, the two levers most cafes have direct control over and rarely touch after opening week.

Food cost is the other warning light worth checking. For a cafe, 25% to 28% is the healthy range. Past 35%, something's wrong, and it's worth finding out what before your next quarterly review rather than after it.

Coffee is carrying more than you think

Coffee typically runs gross margins of 65% to 75%, the highest of anything on a cafe menu. That means every flat white and long black you sell is quietly subsidising the items on your board that don't perform nearly as well. Which is fine, that's normal in any menu. The problem is most operators never actually check which items are doing the subsidising and which are the ones being subsidised.

The labour question most menus ignore

Ingredient cost is only half the picture. The other half is time, and it rarely makes it into the pricing conversation. A drink that sells for $6.50 and takes ninety seconds to make can outperform one that sells for $8.50 and takes four minutes, once you account for the labour sitting inside each cup. On a busy Saturday, that difference compounds fast. It's not about cutting the slow drink. It's about knowing it's slow, and pricing or positioning it with that in mind.

This is exactly why batch prepared drinks carry such strong labour adjusted margin. Make a cold brew concentrate once and pour it to order all week, and the per cup labour after that first batch is close to zero. It's part of why we built Cold Wolff the way we did, a concentrate a cafe can prep once and use across a full week of service rather than making cold coffee to order every single time.

Why this needs checking more than once a year

Most menus get repriced annually if they get touched at all. With food costs moving as much as they have, that's no longer often enough. A coffee or dairy cost that shifted six months ago and never got recosted is margin you've already lost without noticing, because nothing about the menu looks any different from the outside.

What to actually do this week

Pull your POS report and sort by item, not by total revenue but by contribution once you factor in ingredient cost and rough labour time per cup. You're looking for two things. The items quietly carrying the menu, which deserve to stay visible and pushed. And the items that look busy on the till roll but barely clear their own cost once labour's counted, which need a price adjustment, a faster method, or an honest conversation about whether they stay on the board at all.

Do this quarterly, not annually. The gap between the two is real money.

Pull your last quarter's numbers this week. What's the one item on your menu you'd bet is losing you money once labour's counted in, and have you actually checked?