There is probably a dish on your menu that is quietly losing money. It has been there since opening, the kitchen likes making it, and a handful of regulars order it. Nobody has checked whether it earns more than it costs to serve, because nobody has measured. And next to it, a dish that could be paying your rent sits in the middle of the second column, where nobody looks, described in four words that give a guest no reason to want it.

That is what a lot of menus look like: dishes arranged by feel, prices set by instinct, and nobody checking them against the numbers afterwards. A menu like that works against the kitchen. It steers guests toward low-margin dishes and hides the ones that make money, and the owner doesn't see it, because nobody has worked it out.

Menu engineering turns this around. Each dish is measured on two things, how much it earns and how often it sells, and decisions are made from the numbers rather than from gut feeling. Michael Kasavana and Donald Smith formalized it in the 1980s, restaurateurs have refined it for four decades, and as far as I can see it is still rarely used.

A menu is a financial document with a creative face. Without the numbers you lose money; without the creativity you lose the soul. You have to keep both.

01 — The Matrix

Four Quadrants, Four Decisions

Each dish gets a place on a grid with two axes. The vertical axis is contribution margin: how much money the dish brings in once its food cost is taken off. The horizontal axis is popularity: how often it is ordered compared with the other dishes. The axes split the grid into four quadrants, and each one calls for something different.

• Stars ★ — High margin, high popularity. They sell well and make good money. Protect them and feature them, and change them only when the numbers force you to. A dish that costs €5.50 to make and sells for €22 brings in €16.50 a plate, and at high volume that is what pays the rent. When you find a Star, the main job is not to break it.
• Puzzles ? — High margin, low popularity. They make money but don't sell enough. To me this is the most interesting quadrant, because the fix is usually not in the kitchen but in how the dish is presented. It doesn't sell because the description is flat, or it is buried on the page, or the server never mentions it, or the guest doesn't know the name. When a Puzzle becomes a Star through better placement, margin grows with no recipe change, no extra food cost and no extra work.
• Plowhorses ⏚ — Low margin, high popularity. Guests love them, but each order brings in less than it could. A typical case: a generous pasta with expensive imported cheese, at €14, ordered by a quarter of your guests. At 35% food cost each plate earns €9.10, while two lines above it sits a dish that costs less and earns more, and nobody looks at it. The fix is small and precise: 15 grams less cheese, one ingredient swapped for something guests won't notice, or the price up by €1. Small changes, and you see them in the numbers.
• Dogs ✗ — Low margin, low popularity. Few people order them and they make little money. Yet they take up room: on the page, in the guest's head while choosing, in prep, on the shelf in the walk-in. Each Dog holds a place where a Star could be. Take it off. Or, if you believe in the idea, rebuild it from scratch: new recipe, new price, new name, new place on the page. Just don't keep it on the menu out of sentiment.

02 — The Arithmetic

Three Numbers Are Enough

For each dish you need three numbers: food cost per portion, selling price and how many portions sell per week. Everything else comes from those.

• Contribution Margin = Selling Price − Food Cost. A dish priced at €22 with a food cost of €5.50 has a CM of €16.50. This is the money the dish actually brings in, what is left for wages, rent and everything else.
• Food Cost Percentage = (Food Cost ÷ Selling Price) × 100. The same dish: 25%. Below 30% is generally healthy; above 35% is a warning sign.
• Popularity Share = (Units sold of this dish ÷ Total units sold across all dishes) × 100. If you sold 40 of the lamb out of 200 total mains that week, the lamb's popularity share is 20%.

To sort the dishes you need two lines. The first is the average contribution margin: add up every dish's CM and divide by the number of dishes. The second is the popularity threshold: 100% divided by the number of dishes, times 0.7. The 0.7 allows for natural variation: a dish counts as “popular” if it gets at least 70% of the orders it would get if every dish sold equally.

Plot each dish: CM on the vertical axis, popularity on the horizontal. Draw lines at the average margin and at the threshold. You get four quadrants: Stars top right, Puzzles top left, Plowhorses bottom right, Dogs bottom left. There's no interpretation here, just arithmetic, and with a spreadsheet and your POS data it takes about two hours.

03 — An Example

Five Dishes, One Week

Take a restaurant with five main courses and one week of service:

• Grilled Sea Bass: Food cost €7.20 · Price €24 · CM €16.80 · Sold 45 → 25% popularity. ★ Star.
• Braised Lamb Shoulder: Food cost €5.50 · Price €22 · CM €16.50 · Sold 38 → 21% popularity. ★ Star.
• Roasted Celeriac: Food cost €3.10 · Price €16 · CM €12.90 · Sold 18 → 10% popularity. ? Puzzle.
• Risotto Nero: Food cost €4.85 · Price €14 · CM €9.15 · Sold 52 → 29% popularity. ⏚ Plowhorse.
• Burger: Food cost €4.60 · Price €13 · CM €8.40 · Sold 27 → 15% popularity. ✗ Dog.

Average CM: €12.75. Popularity threshold: (100 ÷ 5) × 0.7 = 14%.

Look at the risotto. It is the most popular dish on the menu, ordered over a third more often than the lamb. But each risotto brings in €7.35 less margin than each lamb. Over one week: 52 risottos × €9.15 = €475.80, against 38 lambs × €16.50 = €627.00. Fewer orders, more money. If a better layout moves just ten risotto orders a week to the lamb, that is €73.50 a week, roughly €3,800 a year in extra margin. No new recipe, no price increase, no extra work, just a different arrangement of words on the page.

And the burger sits in the bottom-left quadrant: low popularity, low margin, taking up a place a future Star could fill. Take it off and put on something that has a chance to earn.

04 — The Page

Where Guests Look on the Page

A menu is not a neutral surface. Eye-tracking studies have tried for years to pin down where guests look and where they don't. The layout tells the guest something, whether the designer meant it to or not. Better that you decide what than leave it to chance.

• The hotspot. By the usual rule, on a two-panel menu the top right of the right panel gets the most attention; on a single page, the top center. That is the place for your highest-margin dish. Put a Plowhorse there and you have given your most valuable spot to a dish that can't pay for it.
• First and last in a section. The first dish in a section and the last one before the next get more attention than the rest. The guest reads the first to get their bearings, then looks at the last before deciding. Stars go first. Puzzles go last, where the final glance might turn them into an order.
• Setting a dish apart. A dish set apart by a light box, a different typeface, a “Chef's Selection” note or just more white space around it draws more attention than its neighbors; 30–40% more is the figure often quoted. It is a tool for turning Puzzles into Stars, so use it for that.
• The anchor. The most expensive dish in a section changes how the guest sees everything below it. A €32 fish special at the top makes the €22 lamb look like a sensible choice, and the lamb has the higher margin. The anchor isn't there to sell; it is there to make everything else easier to buy.

05 — The Price

How a Number Feels

Pricing isn't only math; it's also psychology. The same number, written differently, lands differently with the guest. There is a fair amount of hospitality research on this.

• Drop the currency symbol. “22” makes guests less price-sensitive than “€22”, and “€22” less than “€22.00”. Cornell University's Center for Hospitality Research measured a statistically significant rise in average spend when currency signs were removed. The symbol reminds the guest they are spending money. The number on its own reads as part of the dish, not as a cost.
• No price column. When prices are lined up in a right-hand column, guests run their eye down it and compare, then pick the cheapest dish or the second cheapest (so as not to look stingy). When the price sits at the end of the description in the same font and size, it reads as the end of a sentence, not a lone number in a table.
• Put the price inside the description. “Braised lamb shoulder, quince, walnut, smoked yogurt... 22” reads as one thought in which 22 is the natural ending, not a hurdle.
• No round numbers for the pricier dishes. €21 or €23 feels more considered than €20 or €25. A round number suggests the price was picked out of the air; an uneven one suggests it was worked out. The effect is small per dish, but it adds up across the menu.

06 — The Words

A Description Is a Small Ad

Most menu descriptions are ingredient lists. “Lamb shoulder, quince, walnut, smoked yogurt.” Four nouns and three commas. The guest knows what is on the plate but has no reason to want it: no texture, no temperature, no story, nothing to look forward to. It's like selling a house by listing the materials: “Brick, glass, wood, concrete.”

A good description is a small ad that makes the guest want the dish in under 25 words. For example:

“Slow-braised lamb shoulder — eight hours, falling apart — with caramelized quince, toasted walnut, and a smoked yogurt that cuts through everything.”

A few rules:

• Lead with the method or the sensation, not the main ingredient. “Slow-braised” says comfort, craft and time invested straight away. “Lamb shoulder” is just the raw material. The method carries the story; the main ingredient is a detail.
• One sensory word. “Falling apart.” “Crispy.” “Silky.” One brings the dish to life. Two are fine. Three are a perfume catalogue.
• Origin or technique, when there's a reason. “Island-foraged herbs” or “72-hour fermented” says care and intention. “Served with seasonal vegetables” says nothing. It is filler, taking the place of a better word.
• Under 25 words. A description that runs to three lines gets skipped. Not skimmed, skipped. If you can't make the dish appealing in one breath, the description needs cutting, not adding to.

07 — The Habit

Thirty Minutes a Month

Menu engineering isn't a one-off job. It is a monthly habit, and it pays more the longer you keep it up. A menu that is sorted out once and then left for a year quietly drifts back to where it started. Supplier prices move, guests change their habits, the season changes what you can get. The matrix has to be recalculated and the decisions made again.

Each month:

• Pull the POS data. Portions sold per dish and revenue per dish. A full month at least, because over shorter periods holidays, weather and chance distort the numbers.
• Update food costs. Check the latest invoices against your costing sheet. If one key ingredient goes up 10%, a Star can slide into Plowhorse territory without anyone noticing until the margin report arrives.
• Replot the matrix. Have dishes moved? A former Star now sitting with the Plowhorses needs attention right away. A Puzzle that is picking up might deserve a better spot on the page.
• Make three decisions. Promote one Puzzle, fix one Plowhorse, decide on one Dog. Three things, done now. Not five, and not “when we have time”.
• Check again in 30 days. Is the Puzzle selling more? Has the Plowhorse margin improved? Is the Dog's replacement working? The answer is in the numbers, not in what the kitchen thinks.

If your costing sheet is up to date and your POS exports are clean, all of this takes about 30 minutes. Those thirty minutes can shift €200–500 a month in margin. Over a year that is €2,400–6,000 for a single restaurant, often enough to cover a slow month or a broken piece of equipment.

There are also mistakes in the design of the menu itself that even a good matrix can't fix:

• Too many dishes. Every dish you add takes a little attention from all the others. A focused menu of 6–8 mains usually sells better than a sprawling one with 15, and it is cheaper to run, easier to keep consistent and quicker for the guest to read. Decision fatigue is real. With fewer choices, the ones that are left sell better.
• No dish in the lead. If every dish carries the same weight on the page, nothing stands out. A menu needs one dish in focus, one that the layout, the description and the server all point to. If you don't choose it, the guest will fall back on the cheapest or the most familiar.
• Servers who haven't been briefed. Even a well-built menu can be undone by a server who recommends the Plowhorse every night because “everyone loves it”. Tell the floor team which dishes to suggest and why. A server's recommendation is probably the strongest tool the menu has, and it costs nothing to point it in the right direction.
• The same menu all year. A menu made for November looks wrong in July. The season changes not only the ingredients but how the menu looks and feels in the hand: the colors, the type, the weight of the paper. A heavy linen menu with burgundy type in August doesn't fit, and guests feel it even if they couldn't say why.

The menu is the piece of paper every guest reads. It is where your kitchen meets your revenue. When it is laid out by feel and priced by instinct, it leaves money on the table, quietly, every service, and that adds up to thousands of euros the kitchen earned and the page lost.

Menu engineering shows you what you otherwise can't see. It turns gut feeling into arithmetic and arithmetic into decisions. It doesn't need expensive software or a consulting degree. It needs a spreadsheet, your POS data, two hours of honest work and a willingness to let the numbers overrule your gut.

If you have the data but not the method, or the method but not the time, that is what I do at AsketCuisine.

Verovatno i na vašem jelovniku postoji jelo koje tiho gubi novac. Tu je od otvaranja, kuhinja ga rado pravi, a naručuje ga nekoliko stalnih gostiju. Niko nije proverio da li zarađuje više nego što košta, jer niko nije merio. A pored njega, jelo koje bi moglo da plaća kiriju stoji u sredini druge kolone, gde se ne gleda, opisano sa četiri reči koje gostu ne daju razlog da ga poželi.

Tako izgleda mnogo jelovnika: jela poređana po osećaju, cene određene od oka, i niko ih posle nije proverio brojevima. Takav jelovnik radi protiv kuhinje. Gosta vodi ka jelima sa malom maržom, a sakriva ona koja zarađuju, i vlasnik to ne vidi, jer to niko nije izračunao.

Analiza jelovnika (na engleskom menu engineering) to okreće. Svako jelo se meri po dve stvari, koliko zarađuje i koliko se prodaje, i odluke se donose po brojevima, a ne po osećaju. Metodu su uobličili Michael Kasavana i Donald Smith osamdesetih, ugostitelji je doteruju već četrdeset godina, a koliko vidim, i dalje se retko koristi.

Jelovnik je finansijski dokument sa kreativnim licem. Bez računa gubite novac, bez kreativnosti gubite dušu. Treba držati oboje.

01 — Matrica

Četiri kvadranta, četiri odluke

Svako jelo dobija mesto na mreži sa dve ose. Uspravna osa je kontribuciona marža: koliko novca jelo donese kad se oduzme trošak hrane. Vodoravna osa je popularnost: koliko se često naručuje u odnosu na ostala jela. Ose dele mrežu na četiri kvadranta, i u svakom se radi nešto drugo.

• Zvezde ★ — Visoka marža, visoka popularnost. Dobro se prodaju i dobro zarađuju. Čuvajte ih i ističite, a menjajte ih samo kad vas brojevi na to nateraju. Jelo koje košta €5.50, a prodaje se za €22, donosi €16.50 po tanjiru, i kad se proda mnogo, to plaća kiriju. Kad nađete Zvezdu, glavni posao je da je ne pokvarite.
• Zagonetke ? — Visoka marža, niska popularnost. Zarađuju, ali se slabo prodaju. Meni je ovo najzanimljiviji kvadrant, jer popravka obično nije u kuhinji, nego u tome kako se jelo predstavlja. Ne prodaje se jer je opis bled, ili je jelo zatrpano na stranici, ili ga konobar ne pominje, ili gost ne zna šta je to. Kad Zagonetka, bolje postavljena, postane Zvezda, marža raste bez ikakve promene recepta, bez dodatnog troška hrane i bez dodatnog rada.
• Radni konji ⏚ — Niska marža, visoka popularnost. Gosti ih vole, ali svaka narudžba donosi manje nego što bi mogla. Tipičan primer: izdašna pasta sa skupim uvoznim sirom, po €14, naručuje je četvrtina gostiju. Na 35% troška hrane, tanjir donosi €9.10, a dva reda iznad stoji jelo koje manje košta i više zarađuje, i niko ga ne gleda. Popravka je mala i precizna: 15 grama sira manje, jedan sastojak zamenjen nečim što gost neće primetiti, ili cena €1 veća. Male izmene, a vidi se u brojevima.
• Psi ✗ — Niska marža, niska popularnost. Malo ko ih naručuje i malo zarađuju. A zauzimaju mesto: na stranici, u glavi gosta dok bira, u pripremi i na polici u hladnjači. Svaki Pas drži mesto na kome bi mogla da bude Zvezda. Skinite ga. Ili, ako verujete u ideju, napravite ga iznova: nov recept, nova cena, novo ime, novo mesto na stranici. Samo ga nemojte držati na jelovniku iz sentimentalnosti.

02 — Aritmetika

Tri broja su dovoljna

Za svako jelo trebaju vam tri broja: trošak hrane po porciji, prodajna cena i koliko se porcija proda nedeljno. Iz njih se vidi sve ostalo.

• Kontribuciona marža = Prodajna cena − Trošak hrane. Jelo sa cenom od €22 i troškom hrane od €5.50 ima KM od €16.50. To je novac koji jelo stvarno donosi, ono što ostaje za plate, kiriju i sve ostalo.
• Procenat troška hrane = (Trošak hrane ÷ Prodajna cena) × 100. Isto jelo: 25%. Ispod 30% je uglavnom zdravo, a iznad 35% je znak upozorenja.
• Udeo popularnosti = (Prodatih jedinica ovog jela ÷ Ukupno prodatih svih jela) × 100. Ako ste prodali 40 porcija jagnjetine od ukupno 200 glavnih jela te nedelje, udeo popularnosti jagnjeta je 20%.

Da biste jela razvrstali, trebaju vam dve linije. Prva je prosečna kontribuciona marža: saberete KM svih jela i podelite brojem jela. Druga je prag popularnosti: 100% podeljeno brojem jela, pa puta 0.7. Faktor 0.7 ostavlja mesta za prirodna kolebanja: jelo je „popularno“ ako dobije bar 70% narudžbi koje bi mu pripale kad bi se sva jela prodavala podjednako.

Ucrtajte svako jelo: KM na uspravnoj osi, popularnost na vodoravnoj. Povucite linije na prosečnoj marži i na pragu. Dobijete četiri kvadranta: Zvezde gore desno, Zagonetke gore levo, Radni konji dole desno, Psi dole levo. Tu nema tumačenja, samo računa, i sa tabelom i podacima iz kase traje otprilike dva sata.

03 — Primer

Pet jela, jedna nedelja

Uzmimo restoran sa pet glavnih jela i jednu nedelju rada:

• Grilovani brancin: Trošak hrane €7.20 · Cena €24 · KM €16.80 · Prodato 45 → 25% popularnosti. ★ Zvezda.
• Dinstana jagnjeća plećka: Trošak hrane €5.50 · Cena €22 · KM €16.50 · Prodato 38 → 21% popularnosti. ★ Zvezda.
• Pečeni celer: Trošak hrane €3.10 · Cena €16 · KM €12.90 · Prodato 18 → 10% popularnosti. ? Zagonetka.
• Rižoto Nero: Trošak hrane €4.85 · Cena €14 · KM €9.15 · Prodato 52 → 29% popularnosti. ⏚ Radni konj.
• Burger: Trošak hrane €4.60 · Cena €13 · KM €8.40 · Prodato 27 → 15% popularnosti. ✗ Pas.

Prosečna KM: €12.75. Prag popularnosti: (100 ÷ 5) × 0.7 = 14%.

Pogledajte rižoto. Najpopularnije je jelo na jelovniku, naručuje se za više od trećine češće od jagnjetine. Ali svaki rižoto donosi €7.35 manje marže od jagnjetine. Za nedelju dana: 52 rižota × €9.15 = €475.80, naspram 38 porcija jagnjetine × €16.50 = €627.00. Manje narudžbi, a više novca. Ako bolji raspored na jelovniku prebaci samo deset narudžbi nedeljno sa rižota na jagnjetinu, to je €73.50 nedeljno, oko €3.800 godišnje veće marže. Bez novog recepta, bez većih cena i bez dodatnog rada, samo drugačiji raspored reči na stranici.

A burger je u donjem levom kvadrantu: mala popularnost, mala marža, i zauzima mesto koje bi mogla da popuni neka buduća Zvezda. Skinite ga i stavite nešto što ima šanse da zaradi.

04 — Stranica

Gde gost gleda na stranici

Jelovnik nije neutralna površina. Istraživanja praćenja pogleda godinama pokušavaju da utvrde gde gosti gledaju, a gde ne. Raspored na stranici gostu nešto govori, hteo to dizajner ili ne. Bolje je da to vi odredite nego da prepustite slučaju.

• Najgledanije mesto. Po uobičajenom pravilu, na jelovniku na dve strane najviše pažnje dobija gornji desni deo desne strane, a na jelovniku na jednoj strani gornja sredina. Tu ide jelo sa najvećom maržom. Ako tu stavite Radnog konja, najskuplje mesto na stranici dali ste jelu koje ne može da ga plati.
• Prvo i poslednje u grupi. Prvo jelo u grupi i poslednje pre sledeće grupe dobijaju više pažnje od ostalih. Gost pročita prvo da se snađe, pa pređe na poslednje pre nego što odluči. Zvezde idu prve. Zagonetke idu poslednje, gde ih poslednji pogled može pretvoriti u narudžbu.
• Izdvojeno jelo. Jelo izdvojeno nenametljivim okvirom, drugim fontom, oznakom „Izbor kuhinje“ ili samo sa više praznog prostora oko sebe privuče više pažnje od jela oko njega; često se navodi 30–40% više. To je alat za pretvaranje Zagonetki u Zvezde, i za to ga i koristite.
• Sidro. Najskuplje jelo u grupi menja to kako gost vidi sve ispod njega. Riba od €32 na vrhu liste učini da jagnjetina od €22 deluje kao razuman izbor, a jagnjetina pritom ima veću maržu. Sidro nije tu da se proda, nego da se sve ostalo lakše prodaje.

05 — Cena

Kako broj deluje

Cena nije samo matematika, nego i psihologija. Isti broj, napisan drugačije, gost doživi drugačije. O tome postoji dosta ugostiteljskih istraživanja.

• Bez oznake valute. Na „22“ gost manje reaguje nego na „€22“, a na „€22“ manje nego na „€22.00“. Centar za istraživanja u ugostiteljstvu Univerziteta Cornell izmerio je statistički značajan rast prosečne potrošnje kad su oznake valute uklonjene. Oznaka podseća gosta da troši novac. Sam broj doživi se kao deo jela, a ne kao trošak.
• Bez kolone cena. Kad su cene poravnate u desnoj koloni, gost prelazi pogledom niz kolonu i poredi, pa bira najjeftinije ili drugo najjeftinije jelo (da ne ispadne škrt). Kad cena stoji na kraju opisa, istim fontom i veličinom, čita se kao kraj rečenice, a ne kao usamljen broj u tabeli.
• Cena u opisu. „Dinstana jagnjeća plećka, dunja, orah, dimljeni jogurt... 22“ čita se kao jedna misao, u kojoj je 22 prirodan kraj, a ne prepreka.
• Skuplja jela bez okruglih cena. €21 ili €23 deluje promišljenije od €20 ili €25. Okrugao broj ostavlja utisak da je cena izabrana napamet, a neokrugao da je izračunata. Po jelu je to mala razlika, ali se na celom jelovniku sabere.

06 — Reči

Opis je mala reklama

Većina opisa na jelovnicima je spisak sastojaka. „Jagnjeća plećka, dunja, orah, dimljeni jogurt.“ Četiri imenice i tri zareza. Gost zna šta je na tanjiru, ali nema razlog da to poželi: nema teksture, temperature, priče, ničega čemu bi se radovao. Kao da kuću prodajete nabrajanjem materijala: „Cigla, staklo, drvo, beton.“

Dobar opis je mala reklama koja za manje od 25 reči probudi želju. Na primer:

„Spora jagnjeća plećka — osam sati, raspada se — sa karamelizovanom dunjom, tostiranim orahom i dimljenim jogurtom koji sve preseca.“

Nekoliko pravila:

• Počnite od načina pripreme ili osećaja, a ne od glavne namirnice. „Spora“ odmah govori o toplini, zanatu i uloženom vremenu. „Jagnjeća plećka“ je samo sirovina. Priču nosi način pripreme, a glavna namirnica je detalj.
• Jedna reč za osećaj. „Raspada se.“ „Hrskava.“ „Svilenkasta.“ Jedna dočara jelo. Dve su u redu. Tri su već katalog parfema.
• Poreklo ili tehnika, kad za to ima razloga. „Bilje sa ostrva“ ili „fermentisano 72 sata“ govori o brizi i nameri. „Servirano sa sezonskim povrćem“ ne govori ništa. To je popuna na mestu gde bi mogla da stoji bolja reč.
• Manje od 25 reči. Opis od tri reda gost preskoči. Ne preleti, nego preskoči. Ako jelo ne možete da opišete u jednom dahu, opis treba skratiti, a ne dopuniti.

07 — Navika

Trideset minuta mesečno

Analiza jelovnika nije posao koji se uradi jednom. To je mesečna navika, i vremenom se sve više isplati. Jelovnik koji se jednom sredi pa ostavi godinu dana tiho se vrati tamo odakle je krenuo. Cene dobavljača se menjaju, gosti menjaju navike, sezona menja robu. Matricu treba ponovo izračunati i odluke doneti iznova.

Svakog meseca:

• Izvucite podatke iz kase. Koliko je porcija prodato po jelu i koliki je prihod po jelu. Najmanje ceo mesec, jer u kraćem periodu brojeve kvare praznici, vreme i slučaj.
• Osvežite troškove hrane. Uporedite nove fakture sa kalkulacijom. Ako jedan važan sastojak poskupi 10%, Zvezda može da postane Radni konj, a da to niko ne primeti dok ne stigne izveštaj o marži.
• Ponovo ucrtajte matricu. Da li su se jela pomerila? Zvezda koja je sada među Radnim konjima traži da se reaguje odmah. Zagonetka koja se sve više prodaje možda zaslužuje bolje mesto na stranici.
• Donesite tri odluke. Jednu Zagonetku istaknite, jednog Radnog konja popravite, o jednom Psu odlučite. Tri stvari, i to odmah. Ne pet, i ne „kad budemo imali vremena“.
• Proverite za 30 dana. Prodaje li se Zagonetka više? Je li marža Radnog konja bolja? Radi li ono što je zamenilo Psa? Odgovor je u brojevima, a ne u tome šta kuhinja misli.

Ako je kalkulacija ažurna, a podaci iz kase uredni, sve to traje oko 30 minuta. Tih trideset minuta može da donese €200–500 mesečno veće marže. Za godinu dana to je €2.400–6.000 za jedan restoran, što je često dovoljno da se pokrije slab mesec ili pokvaren aparat.

Ima i grešaka u samom izgledu jelovnika koje ni dobra matrica ne može da popravi:

• Previše jela. Svako novo jelo uzima malo pažnje od svih ostalih. Sažet jelovnik od 6–8 glavnih jela obično prodaje bolje od razvučenog sa 15, a i jeftiniji je za rad, lakše se drži ujednačenim i gost ga brže pročita. Umor od biranja postoji. Kad je izbor manji, ono što ostane bolje se prodaje.
• Nijedno jelo ne vodi. Ako sva jela na stranici izgledaju jednako važna, ništa se ne ističe. Jelovniku treba jedno jelo u centru pažnje, na koje upućuju i raspored, i opis, i konobar. Ako ga vi ne izaberete, gost će uzeti najjeftinije ili najpoznatije.
• Konobari koji nisu upućeni. I dobro složen jelovnik pokvari konobar koji svako veče preporučuje Radnog konja jer „to svi vole“. Recite sali koja jela da predlaže i zašto. Preporuka konobara je verovatno najjači alat koji jelovnik ima, a ništa ne košta da se usmeri.
• Isti jelovnik cele godine. Jelovnik napravljen za novembar u julu deluje pogrešno. Sezona ne menja samo namirnice, nego i to kako jelovnik izgleda i kakav je u ruci: boje, slova, težinu papira. Težak laneni jelovnik sa bordo slovima u avgustu ne pristaje, i gost to oseti, iako ne bi umeo da kaže zašto.

Jelovnik je papir koji pročita svaki gost. Na njemu se sreću kuhinja i prihod. Kad je složen po osećaju, a cene određene od oka, novac ostaje na stolu, tiho, na svakom servisu, i to se skupi u hiljade evra koje je kuhinja zaradila, a stranica izgubila.

Analiza jelovnika pokazuje ono što se inače ne vidi. Osećaj pretvara u račun, a račun u odluke. Za to ne treba skup program ni konsultantska diploma. Treba tabela, podaci iz kase, dva sata poštenog rada i spremnost da pustite brojeve da pobede osećaj.

Ako imate podatke, a ne znate kako da ih složite, ili znate, ali nemate vremena, time se bavim u AsketCuisine.