Chapter 12 — Key Takeaways

The core claims

  1. You bank dollars, not percentages. Contribution margin — menu price minus plate cost — is what pays labor, rent, and debt. Food cost percentage is the right control number for kitchen execution and the wrong number for comparing two dishes, because it throws away the size of the price tag.
  2. Bellwether proves it on its own menu. The worst food cost on the page (ember trout, 37.1%) out-earns the second-best (Bellwether burger, 24.3%) by $1.70 a plate. Manage by percentage and you cut the trout and push the burger.
  3. A menu is a portfolio with two axes: what a plate earns, and how often guests choose it. Two lines make four quadrants and four default actions.
  4. The margin axis is the weighted average, not the simple one — total contribution margin divided by total units.
  5. The popularity threshold is a convention, not a finding. The conventional 70% rule is defensible; it is not evidence. Move it to 100% of average and two of Bellwether's five items change quadrant.
  6. Both axes are relative to your own menu, so every menu produces Stars and Dogs by construction. Cut the bottom item and a new bottom item appears. Follow the advice literally three times at Bellwether and the Hearth Chicken clears the line by under two cents.
  7. The matrix cannot see labor, station load and throughput, cross-utilization, guest counts as distinct from item counts, veto power, substitution, seasonality, sample size, or beverage. Use it to rank questions. Never let it hand down a verdict.
  8. Sell before you cost, cost before you price, price before you portion, portion before you cut. Most operators start at the expensive end.

The formulas

  contribution margin (CM)  = menu price - plate cost
  menu mix %                = item units / total category units
  popularity index          = item mix % / (1 / number of items)
  popularity threshold      = 0.70 / number of items          (the convention)
  weighted average CM       = total contribution margin / total units
  break-even on a price rise= old total CM / new CM = units you must still sell
  gap-closing rule of thumb = a $1 CM move closes the gap to the average
                              by about $1 x (1 - the item's mix share)

Bellwether's dinner menu, placed

Item Plate cost Price FC% CM Units Mix Quadrant
Hearth Chicken $8.52 | $29.00 29.4% $20.48 96 29.1% STAR
Bellwether burger $5.10 | $21.00 24.3% $15.90 73 22.1% PLOWHORSE
squash and grains $3.95 | $24.00 16.5% $20.05 62 18.8% STAR
pork chop $10.09 | $31.00 32.5% $20.91 54 16.4% STAR
ember trout $10.40 | $28.00 37.1% $17.60 45 13.6% DOG
Totals $2,447.98** | **$8,739.00 28.0% $6,291.02 330 100.0%

Weighted average CM $19.06 · popularity threshold 14.0% · annualized entrée CM $327,133

Three Stars, one Plowhorse, one Dog, no Puzzle. The trout misses the popularity line by 1.2 units a week — which is inside the sampling noise of an entire season, so it is a judgment call, not a finding.

The numbers worth remembering

Figure Why
**$19.06** | Bellwether's weighted CM; a dollar of movement is $17,160 a year
$5.01 the whole menu's CM spread — a flat portfolio, so mix work is a weak lever
14.0% the popularity line: 0.70 × (1 ÷ 5), or 46.2 units a week
40.0% share of dinner sales the entrée matrix actually covers
±2 points sampling noise on any one share at 330 entrées a week
**$3,544** | annual cost of ignoring a $0.40/lb poultry increase on the Star

Monday morning

You should be able to: pull last week's product-mix report, attach current plate costs, compute contribution margin and menu mix for every item in one category, derive the weighted average and the popularity threshold, place every item, mark anything within two points of a line as unplaced, and walk into a meeting with two proposed changes and a dollar figure attached to each — plus the sentence that names what the analysis could not see.