Costs and stock
Menu engineering: keep the dishes that pay the rent
Your bestseller and your most profitable dish are rarely the same plate. Menu engineering is the exercise that tells you which is which.
Klyra Editorial · Published on 6 August 2026 · 8 min read
Food cost percentage ranks dishes badly: a 22% dish selling twice a week contributes less than a 34% dish selling forty times. Menu engineering ranks by cash margin and volume together.
The four quadrants
- Stars — high margin, high volume: protect them, never discount them
- Plough-horses — low margin, high volume: rework the recipe or lift the price slightly
- Puzzles — high margin, low volume: reposition on the menu or rename them
- Dogs — low margin, low volume: remove, unless they anchor the concept
Get the inputs right first
The exercise is only as good as the recipe costing behind it. Yields after trimming and cooking, current supplier prices, and the real portion served — not the one written down two years ago.
Act, then measure
- Change no more than five dishes at once.
- Keep the price change small — under 8% rarely affects volume.
- Wait four full weeks before judging the result.
- Compare margin per cover, not just total revenue.
Make it repeatable
Run the analysis quarterly and whenever a key supplier price moves more than 10%. Once costing updates automatically from purchase prices, the quarterly review takes an hour instead of a weekend.
FAQ
How often should a menu be re-engineered?
Quarterly is a good cadence, plus an out-of-cycle review whenever a major ingredient price moves sharply.
Should I remove every low-margin dish?
No. Some low-margin items anchor the concept or drive footfall. Remove low-margin dishes that also sell poorly.
Keep reading
Keep reading
Start free
Practical guides on HACCP, food cost, rotas and multi-site management.
Start free