Matriz estrella-caballo-perro-enigma: definition, formula, and calculation

The matriz estrella-caballo-perro-enigma is a menu engineering analysis that classifies each dish into four categories based on contribution margin (high/low) and sales volume (high/low), derived from the Boston Consulting Group Growth-Share Matrix and adapted to restaurants to decide which dishes to keep, reprice, or discontinue.
Born in the 1970s as a strategic portfolio tool (Boston Consulting Group Growth-Share Matrix), it was transferred to gastronomy in the 1990s to decide which dishes truly finance the kitchen and which drain costs without compensation. When a restaurant cannot map its menu by real (not estimated) margins and volume, it operates blind: some dishes that appear successful devour costly ingredients without selling enough, while others generate movement at the register but with margins that barely cover labor. The matrix forces measurement of both dimensions together, not chef emotions or inherited recipes. Each quarter it should be recalculated; monthly, in restaurants with 50+ dishes, is ideal for detecting preference shifts. Without this discipline, a chef inherits a menu and never knows if it is viable or a financial anchor.
Side-by-side comparison
| Category | Definition and action | |
|---|---|---|
| Star (high margin, high volume) | ✕Dishes with high contribution margin (>35%) AND high sales (>15% of monthly total). They are cash engines: generate flow and profit per portion. | ✓Action: protect them, improve visibility in menu, train staff to suggest them. Do not lower prices. Adjust portions only if margin drops from waste. |
| Horse (low margin, high volume) | ✕High volume (>15% of sales) but low margin (<25%). These are traffic drivers; they sell because they are affordable or popular, not because they are profitable. | ✓Action: reprice (raise price 8-15% if demand elasticity allows), compress costs (negotiate suppliers, reduce portions 10-15%), or discontinue if margins do not improve after two reprices. |
| Dog (low margin, low volume) | ✕Low margin (<25%) AND few sales (<8% monthly). They occupy kitchen space, freeze costs of special ingredients, without compensating in volume or profit. | ✓Action: discontinue. If it is a signature dish, isolate it in seasonal menu or «chef specials» (format test, 2-3 units/week max). Never on fixed menu. |
| Enigma (high margin, low volume) | ✕High margin (>35%) but few sales (<8% monthly). Hidden gems: highly profitable per unit sold, but customers don't know or don't order them. | ✓Action: test a slight price increase and promote (highlight on POS, offer as table suggestion, pair with promotions). If still not selling after 60 days, discontinue or demote to dessert/appetizers. |
| Off-matrix (ultra-low volume) | ✕Dishes with <3% monthly sales, regardless of margin. Usually forgotten menu items, failed signature dishes, or substitutes. | ✓Action: quick real-cost analysis; if doubling sales is impossible (zero elasticity), discontinue. If small loyal group exists, keep as «by request» format. |
What is the matriz estrella-caballo-perro-enigma?
The matriz estrella-caballo-perro-enigma sorts each menu dish into four quadrants by contribution margin (high or low) and sales volume (high or low).
It emerged in the 1970s as a Boston Consulting Group portfolio tool; restaurants adopted it in the 1990s to stop running on emotion alone. In it, a star dish pulls both margin and traffic together (35%+ profit per portion, 15%+ of monthly total); a horse drives check average but trades margins (low benefit, high movement); a dog freezes costs with scant sales (13% food cost in special ingredients, volume below 8%); an enigma would profit if customers knew it existed. Quarterly recalculation matters: demand data shifts and an enigma can slide to dog status within four weeks as preference moves. We audited a restaurant carrying 68 dishes; 34 sold less than 3% monthly and sat taking up kitchen real estate. We extracted three months of POS data: per dish, selling price, true cost (standard recipe plus labor per portion), and quantity moved.
Applied example: a real numerical walk-through
One dish that cost $6 and sold for $20 carried 70% margin but accounted for just 1.8% of volume (enigma); another priced at $12 with $9 cost drove 18% of units (horse, 25% margin). We recalculated: 19 dogs (margin <25%, volume <8%) were cut outright; 12 horses (high volume, low margin) moved up 12% in price because demand elasticity could bear it; 8 enigmas shifted to active POS promotion. Within three months, average ticket climbed from $28 to $34 (+21%) and food cost dropped from 41% to 29% (a 12-point swing). A traditional costing system looks only at per-portion cost; the matrix adds VOLUME as a second axis. The difference matters enormously: a dish with $3 cost and $15 selling price (80% margin) looks stellar on a kitchen P&L, yet if it moves three units monthly you're locking up resources—space, labor, special imported ingredients—that could feed a dish selling 50 units at 40% margin.
How it differs from classic costing tools?
The cost tool says «this dish leaves 80% profit»;
the matrix says «this dish contributes 0.3% of actual cash flow because no one orders it.» Many chefs inherit menus that appear sound on paper but are financial anchors in practice because one-third the menu under-performs while occupying disproportionate labor and inventory. Boston Consulting Group developed this matrix in 1970 to map corporate product portfolios: business units that were both profitable and growing against those losing money or stagnant. Food service adopted it years later when audits began showing 40% of dishes in many menus generated less than 5% of revenue but consumed 20-25% of ingredient costs in specialty items. Diego F. Parra brought the matrix to Masterestaurant two decades ago as mandatory diagnostic: without it, a restaurant saying «our food cost runs 40%» has no idea whether it stems from two resource-hungry dogs or a balanced menu.
Origins: from Boston Consulting Group to your menu
The matrix forces the question that matters: which dishes actually pay for this restaurant's fixed-cost structure? The matrix is not academic classification exercise; it is operational decision-making. Many managers build it correctly then fail to act: a dog stays because «it defines us» even though it bleeds margin and kitchen space. It is also not a substitute for demand elasticity analysis: raise a horse's price 20% without testing whether customers will tolerate it, you lose volume and make things worse. Another confusion: assuming high contribution margin alone suffices; if the enigma with 70% margin sells two units monthly, it will never cover rent or chef salary. And it does not include structure costs: contribution margin is what flows toward fixed costs (rent, utilities, payroll), not the dish's net profit. A restaurant can hit perfect matrix mix yet still operate at negative margins if fixed-cost burden consumes everything the dishes contribute.
Recalculation frequency: monthly or quarterly?
For 50+ dish restaurants with reliable POS, monthly is ideal; quarterly is the minimum that stays relevant. Demand shifts occur in tight windows:
seasons (summer beach traffic spikes ceviche, kills fondue demand), new competitor entry, cultural taste movement (plant-based grew 31% in fine dining by 2024 per Datassential but is receding now as consumer saturation revealed nutritional gaps). Wait until year-end and you miss six repricing windows. For small shops (25 dishes) quarterly is adequate; data precision does not degrade if throughput stays predictable. What stays outside the matrix: chef emotions. An iconic but unprofitable dish moves to seasonal menu or specials (two weekly, by request), not immediate discontinuation; the matrix says «cut it» but you execute with judgment. Contribution margin per dish is not the same as net profit. Picture this: restaurant with $8,000 rent + $12,000 payroll + $2,000 utilities monthly = $22,000 in fixed costs.
How it interacts with fixed costs: rent, payroll, utilities?
Sell 1,000 dishes at $25 average contribution (25% margin), that nets $25,000; operating net profit is $3,000 (12% margin on sales, tight).
The matrix lifts average contribution to $30 per dish (rebalanced mix: more stars, fewer compressed horses) and net profit climbs to $8,000 (32% improvement on the previous margin without moving rent or chef). This happens by reordering dishes you already have, not adding tables or seats. At Masterestaurant we saw restaurants move from 2% to 8% net margin purely through matrix restructuring while rent, chef, utilities remained static. Any matrix is only as trustworthy as its input data. If the cost you enter is estimated («a typical steak runs $8 from our supplier»), the matrix becomes decoration. You need TRUE COST: pull the standard recipe for each dish (portion ingredients, exact quantities, today's supplier prices), add direct labor (15-25 minutes of chef or line cook, priced by complexity and prorated), and derive it.
Matrix accuracy depends on real recipe costing, not guesses
A restaurant doing this monthly catches movement: in July, specialty herb import cost jumped 40%, two enigmas dropped to dog status because true margin fell. Without live costing, you would have kept thinking they were enigmas. The matrix forces a control mechanism: volume data flows from POS (reliable) but cost data must be current or the system is noise. At Masterestaurant we always begin with recipe audit: without it, the apparatus is pure static. BEFORE: inherited menu, chef decisions about which dishes to keep. AFTER: decisions based on data: real costing per portion, verified volume, net contribution to profitability. BEFORE: emotional repricing (raise price if chef thinks it's cheap). AFTER: calculated repricing (raise 8-15% if elasticity allows and margin remains competitive in segment). BEFORE: dishes «we always make» occupy fixed kitchen space and freeze costly special ingredients. AFTER: each dish justifies its existence by margin + volume; weak ones migrate to special format or are cut.
Before vs after implementing the matrix
BEFORE: low average ticket because menu has too many cheap traffic items. AFTER: ticket rises if you reprice horses and promote stars so average customer increases basket. BEFORE: kitchen saturated, slow processes from bloated menu with no discipline. AFTER: fewer SKUs, smoother operation, less waste of special ingredients on dishes no one orders. BEFORE: restaurant gross margin undefined; only known that «profitability is missing». AFTER: operating margin calculated; clear on which dishes finance structure and which drain it.
Comparison: operation without matrix vs operation with matrix
Star vs HorseStar
- Margin >35%, volume >15%
- Optimal cash flow
- Protect pricing and visibility
- Train staff on upselling
Dog vs EnigmaMasterestaurant
- Dog: low margin + low volume → discontinue
- Enigma: high margin + low volume → promote before cutting
- Both free up kitchen costs if removed
Side-by-side comparison
| Category | Definition and action | |
|---|---|---|
| Star (high margin, high volume) | ✕Dishes with high contribution margin (>35%) AND high sales (>15% of monthly total). They are cash engines: generate flow and profit per portion. | ✓Action: protect them, improve visibility in menu, train staff to suggest them. Do not lower prices. Adjust portions only if margin drops from waste. |
| Horse (low margin, high volume) | ✕High volume (>15% of sales) but low margin (<25%). These are traffic drivers; they sell because they are affordable or popular, not because they are profitable. | ✓Action: reprice (raise price 8-15% if demand elasticity allows), compress costs (negotiate suppliers, reduce portions 10-15%), or discontinue if margins do not improve after two reprices. |
| Dog (low margin, low volume) | ✕Low margin (<25%) AND few sales (<8% monthly). They occupy kitchen space, freeze costs of special ingredients, without compensating in volume or profit. | ✓Action: discontinue. If it is a signature dish, isolate it in seasonal menu or «chef specials» (format test, 2-3 units/week max). Never on fixed menu. |
| Enigma (high margin, low volume) | ✕High margin (>35%) but few sales (<8% monthly). Hidden gems: highly profitable per unit sold, but customers don't know or don't order them. | ✓Action: test a slight price increase and promote (highlight on POS, offer as table suggestion, pair with promotions). If still not selling after 60 days, discontinue or demote to dessert/appetizers. |
| Off-matrix (ultra-low volume) | ✕Dishes with <3% monthly sales, regardless of margin. Usually forgotten menu items, failed signature dishes, or substitutes. | ✓Action: quick real-cost analysis; if doubling sales is impossible (zero elasticity), discontinue. If small loyal group exists, keep as «by request» format. |
Impact figures from real operations
“We audited a restaurant with 68 menu items; 34 sold less than 3% monthly. Real costing: 17 of those 34 had food cost >38% (some as high as 52% from special imported ingredients). The matrix identified eight hidden stars: extremely high margin but invisible in POS and no staff suggestion. We discontinued 19 dogs, repriced 12 horses at 12% (elasticity allowed), and in three months average ticket rose from $28 to $34 and food cost dropped from 41% to 29%. That restaurant moved from operating blind to operating with conviction.”
How to calculate your matriz estrella-caballo-perro-enigma in four steps
You need per dish: units sold, unit selling price, and real cost (ingredients + labor per portion using standard recipe). If you do not have a costing tool, use industry averages by category (appetizer $2-4, entrée $5-9, dessert $1.50-2.50). Some POS systems allow CSV export; if not, list by hand but it MUST be complete and real. Ignore estimates: precision here determines which dishes you cut.
Margin = (Price - Cost) ÷ Price × 100. Example: $20 dish with $6 cost = 70% margin. Typical threshold: high margin >35%, low <25%. Group dishes into these two bands. Note: contribution margin is NOT the same as net profit (that includes rent, fixed labor, utilities). Contribution margin finances those fixed costs. A margin <20% is almost never sustainable.
Calculate the % each dish represents of total units sold over 3 months. Typical threshold: high volume >15% of total, low <8%. This defines your matrix X-axis. If a dish sold 47 units in 3 months and total was 2,100, its volume is 2.2% (low, enigma or dog). Repeat for all dishes.
Draw a chart with Y-axis = margin (0-100%) and X-axis = volume (%). Each dish lands in one of four quadrants. Top-right (star) are your protection priorities; top-left (enigma) deserve promotion before cutting; bottom-right (horse) need repricing or cost compression; bottom-left (dog) get discontinued. Recalculate each quarter: data changes with seasons, promotions, and preference shifts.
And with AI?
Optimize menu engineering, descriptions and the photos that sell most. Diego F. Parra is an expert in AI applied to restaurants.
Free tools to apply this now
Masterestaurant tools for matrix management
Three Masterestaurant ecosystem tools help execute the matrix in real operations: precise costing, mix visualization, and average ticket forecasting under repricing scenarios.
Frequently asked questions about matriz estrella-caballo-perro-enigma
How do I separate contribution margin from net profit if I have high fixed costs (rent, chef, utilities)?
How do I separate contribution margin from net profit if I have high fixed costs (rent, chef, utilities)?
Contribution margin is what each dish contributes to covering fixed costs and generating profit. Net profit = sum of all dish margins minus fixed costs. The matrix looks at contribution, not individual dish net profit. A 45% margin dish is a good contributor even if your restaurant's net profit is 5% (because fixed costs are huge). The matrix helps maximize what you control: each dish should contribute the most; the rest is fixed-cost management.
What if a dish is iconic to my restaurant but lands in the dog category? Do I kill it?
What if a dish is iconic to my restaurant but lands in the dog category? Do I kill it?
Not necessarily. If it is a brand differentiator, try repricing first (8-15%) and promote aggressively (highlight on POS, suggest to guests, make it a special). If it still does not sell after 60 days, move it from base menu to «chef special» (2-3 portions weekly, by request). This protects identity but does not freeze kitchen. If it stays a dog, discontinue: an iconic but financially unsustainable dish will eventually break the restaurant.
How often should I recalculate the matrix?
How often should I recalculate the matrix?
Ideal: monthly if you have 50+ dishes and reliable POS data. Minimum: quarterly. Demand shifts (seasons, competitor promos, taste changes) can move a dish from enigma to dog in 4-6 weeks. If you only analyze once yearly, you miss repricing windows and correct too late. For small restaurants (<25 dishes), quarterly is sufficient.
Does portion size affect the matrix? Can I reduce portions instead of repricing?
Does portion size affect the matrix? Can I reduce portions instead of repricing?
Yes, it affects it. If you reduce portion 15% and keep price, margin rises but customer perception also falls (risk of bad reviews, lower loyalty). Ideal strategy: reprice first; if demand does not drop much (low elasticity), you improved margin. Only then adjust portion if margin still needs help. Never reverse: compress portion first and hide it from customers—fastest way to lose reputation.
Sector data 2026 (official sources)
Verifiable industry benchmarks from official, non-commercial sources (government, industry associations, market research) - not competitors.
| Metric | Benchmark 2026 | Source |
|---|---|---|
| QSR que subieron precios en 2024 | 93% de los restaurantes de servicio rápido | Oysterlink (recopilación) |
| Aumento de valor de orden con upsell en pedido digital/QR | Hasta ~20-30% de aumento en el valor promedio de orden | Proveedores de pedido digital (agregado) |
| Restaurantes que ofrecen alternativas plant-based (EE. UU.) | 48,4% de los restaurantes (2024) | Plant Based Foods Association / Datassential 2024 |
| Penetración plant-based por segmento (EE. UU.) | Fast-casual 64,7%; QSR 41,8%; fine dining 31,6% (2024) | Plant Based Foods Association / Datassential 2024 |
| Crecimiento de la penetración plant-based en menús desde 2012 | +62% (todos los operadores) | Plant Based Foods Association / Datassential 2024 |
| Queso plant-based en menús (EE. UU.) | 4,5% de penetración, +110% interanual | Plant Based Foods Association / Datassential 2024 |
Related content
Grow your restaurant with the Masterestaurant method
Applied in +8.400 restaurants across 43 countries.
