Stars, plowhorses, puzzles, dogs

Stars — protect them

High popularity, high contribution. The trap is complacency: stars drift. A star whose protein has crept up 20% since you costed it is quietly the most expensive problem on the menu, because it sells the most. Recost stars first, every time.

Plowhorses — reprice or rework

High popularity, low contribution. These carry your covers and not your profit. Because demand is proven, a modest price increase is unusually safe here. The alternative is cost reduction — a cheaper garnish, a better yield, a smaller portion of the most expensive component.

Puzzles — reposition before you cut

Low popularity, high contribution. Something is stopping people ordering a dish that makes good money. Usually it's placement, a flat description, or a price that reads high next to its neighbours. Move it, rewrite it, or have servers mention it for two weeks before concluding it doesn't work.

Dogs — cut, with two exceptions

Low popularity, low contribution. Cut them: they consume prep, inventory SKUs, and menu space. The exceptions are dishes that exist for a reason other than their own P&L — the vegan entrée that lets a table of six book, the kid's plate, the dish a group of regulars comes in for.