Dish photography and menu descriptions: what YOUR menu needs before you spend a dollar

What works better for an independent under 40 seats: photos or copy?

An independent restaurant under 40 tables should rewrite the descriptions of the eight dishes carrying its sales mix before hiring any photo shoot, because it costs nothing in cash and takes one afternoon.

The Guadalajara case we reviewed at Masterestaurant says it plainly: 28 tables, a 1,400 USD quote to photograph 46 dishes, and a mix where 71 % of revenue came from nine items, seven of which carried four-word descriptions along the lines of «grilled chicken». We rewrote those seven with sourcing of the main ingredient, cooking technique and a named side, touching no prices, and the average check climbed from 18.40 to 20.10 USD across eleven weeks. The 1,400 USD stayed in the account. The menu, meanwhile, is already working. Two different currencies, spent at different moments of the guest's journey, and the industry confuses them with a cheerfulness that still surprises me.

Photography buys attention; copy buys permission to charge

Photography works BEFORE the guest picks the venue —on the delivery listing, on the map, in the feed— and there it is unbeatable, because nobody smells anything through a screen and the image IS the product. Copy works inside, with the guest seated and the server standing, while they decide between two dishes that may carry 27 points of margin difference between them. If your business bills 80 % in the dining room, you are buying attention you already had. And that duplicated attention, in practice, never shows up in Tuesday's register. Every menu change is judged by its effect on mix-weighted contribution margin, never by how pretty the page turned out. Say a rewritten description lifts by two points the share of the dish carrying 68 % margin and drops by two the share of the 41 % dish: you gained roughly 0.54 points of overall margin even if the average check never moves a cent.

The financial frame almost nobody applies: mix-weighted margin

That calculation demands standard recipes and per-portion costing closed before you touch anything, with plate food cost inside the range the National Restaurant Association calls optimal, 28-35 %, running 30-34 % in casual dining. Without those two documents, any menu investment —photo or text— is faith, not management. Faith does not show up on the income statement. If your operation bills more than 60 % through aggregators or a ghost kitchen, put money into photography first and do it properly, because on that channel the image occupies the seat the server occupies in the dining room. A dish without a photo on an app competes against twelve that have one inside the same three-inch screen, and the description gets truncated to two lines before any thumb reads it. The argument flips here: text does little because hardly anyone opens it. Digital spending evidence points the same way —Taco Bell reports 20 % higher spend per order on digital self-service versus a human cashier— and that gap rests on interface and image.

Best for pure delivery and dark kitchens: here photography does rule

Photograph your twelve highest-rotation dishes, not all 46. In high-ticket operations the description is worth more than any photograph, and the photo can even work against you by fixing a visual expectation the plated dish will then have to match exactly. Fine dining runs food cost at 34-40 % according to the National Restaurant Association, meaning tighter gross margin per plate and compensation through price; that price needs narrated justification: provenance, cut, curing time, the producer's name. There is a typographic detail with hard evidence behind it: Cornell measured that removing the dollar sign from the menu raises spend per person by 8.15 %. It costs one afternoon of layout work. I would start there before calling the photographer, especially if your menu rotates by season and images expire every four months. Three scenarios exist where rewriting descriptions will not move your needle, and it deserves saying bluntly.

When NOT to pick the popular option (rewriting descriptions)?

First: a kitchen with no standard recipe and no per-portion costing, because you will push volume toward dishes whose real margin you do not know and may end up selling more of what pays you least.

Second: a flat mix with no concentration, where no single dish clears 6 % of sales; that is a menu-engineering problem with too many items, not a shortage of adjectives. Third: delivery-dominant sales on listings without images, where the text never gets read. I will add a fourth I keep running into: menus with median food cost above the 32.4 % the National Restaurant Association reports for limited service, where the hole sits in purchasing, not in the writing. Four concrete signals tell you somebody is selling you the wrong lever. One: the photographer quotes by dish count rather than by dishes that carry the mix —46 photos when nine generate 71 % of revenue is money turned into dead files.

Red flags when comparing the two options

Two: the proposal promises «higher sales» without naming contribution margin or average check, two figures anyone can audit eight weeks later. Three: the copywriter delivers adjectives with no ingredient, no technique, no origin; «delicious» and «homemade» justify not one cent of price. Four: nobody asks for your menu-engineering matrix before starting, a sign they will decorate dishes at random. If two of these four show up, stop the purchase and ask for the numbers. Pull the last 90 days of sales, sort by revenue and keep the dishes that add up to 70 %; that is usually a short handful, seven to ten even on menus with 40 items. For each one write three verifiable elements: where the main ingredient comes from, how it is cooked, and what accompanies it, named. No «homemade», no «exquisite». Then measure two things at week eight: each dish's share of the mix and the weighted contribution margin; if overall margin rose half a point on monthly revenue of 60,000 USD, that is 300 USD appearing without spending anything.

What I would do this week with the menu you already have?

And with that you have already funded the photo shoot for which you now know exactly which dishes deserve coverage. Photography buys ATTENTION and description buys PERMISSION to charge more.

Two different currencies, spent at different moments of the guest journey: the image works before the restaurant gets chosen, the copy works inside, once the guest is seated and weighing two dishes that sit 27 margin points apart. On the restaurant's books a photo is marketing CAPEX that amortizes badly, since it rarely survives one full menu cycle; rewritten descriptions are a process improvement that outlives price changes and supplier swaps, which is why that is the first lever I pull whenever cash is tight. Demand elasticity responds differently to each. An attractive photo lifts volume on the dish it shows but hands the guest no argument for accepting a higher price; a description carrying provenance and technique shifts perceived value upward, and that shift is what lets you raise a price 8 % without losing units.

Where the two routes genuinely diverge?

Execution risk splits them too:

a mediocre shoot leaves you with images that age badly and that you will publish anyway because you already paid, whereas weak copy gets rewritten the same day you spot it in the mix report, with no sunk cost involved.