Redesigning an old menu: a checklist before you touch a single dish

Most reprints start in the wrong place. Someone opens last year's file, swaps three dishes, nudges two prices and sends the PDF to print. Six weeks later the new card looks newer and still sells the same problems: a starter nobody cooks well, a main that clogs the pass, a dessert that looks like an afterthought. A redesign is not a facelift. It is an audit of what the kitchen can do, what the till already knows, and what the guest can read in twenty seconds under restaurant light. Do the work on paper first. Touch the layout last.

Photograph the card you already have

Before you change a line, photograph every face of the current menu in the dining room, at night, with the lights you actually use. Phone flash lies. You need to see the grease halo on the cover, the crease that no longer sits in the score, the type that disappears against a stained panel, the price column that has been written over in pen. Those pictures are the brief. If the 11 pt descriptions are unreadable at table 12, the next file should not ship the same size.

Keep one clean copy in a sleeve and one worn copy on the desk. The clean copy tells you the intended hierarchy. The worn copy tells you where thumbs land, which panel gets the most traffic, and which section staff have annotated because the printed line is wrong. A redesign that ignores the worn copy will reprint the same friction.

Count what sells, not what you are proud of

Pull four to eight weeks of till data, grouped by dish, inside each section. You want two numbers: units sold and contribution margin in euros (selling price excluding VAT minus ingredient cost, including garnish and the sauce made in a batch). Pride is a bad editor. The sea bass you plated for the photographer may be a dog. The omelette you barely mention may be carrying Tuesday lunch.

Mark each dish as a star, plough horse, puzzle or dog, or simply as keep, push, reprice or cut. The labels matter less than the decision. A redesign that keeps every dish and only changes the typeface is a reprint, not a redesign. Cutting three dogs from a 42-item card is often the single change that lets body type jump from 9 pt to 11 pt and makes the new layout look expensive.

Ask the kitchen what the card is hiding

Sit with the head chef for thirty minutes with the current menu on the pass. For each section ask three questions: which dishes share a pan or a fryer at 20:30, which dishes need a cook who is not always on, and which dishes you would drop tomorrow if the card allowed it. The answers are layout constraints. A section of eight mains that all hit the grill is not a design problem; it is a service problem the new card should refuse to reprint.

• Shared equipment: if four mains need the same plancha, one of them should leave or move to a lunch-only line.
• Skill dependency: a dish that only one cook can finish is a risk, not a signature, until a second person can plate it.
• Prep load: anything that needs a two-day brine or a sauce that splits if held is expensive even when the food cost looks fine.
• Allergen theatre: dishes that generate a five-minute conversation at every table belong on a clearer line, not a footnote.

Write the answers in the margin. The designer, or the person generating the next file, should see them. A beautiful two-column A4 that still sends four steaks and two burgers to one station at 21:00 is a failed redesign.

Fix the legal lines and the tired copy first

In the EU the 14 allergens under Regulation 1169/2011 must be available in writing. If your current card hides them in a 7 pt footer or a QR code that fails, the redesign is the moment to put a readable system on the page: a short key, consistent marks next to dishes, and a line that tells the guest to ask. Do not invent a second language of icons. One system, used everywhere.

Then rewrite the descriptions you are about to typeset. Old menus accumulate adjectives that no longer match the plate: homemade when it is not, seasonal when the garnish is the same in January, a farm name you stopped buying from. One sensory sentence is enough for most dishes. Save origin for the one ingredient that is actually specific. Prices should be decided now, not after the first proof, so the column width is honest and you do not end up squeezing 24,50 into a slot built for 18.

Choose format and paper before you choose a style

An old A4 bifold that has grown to 60 lines per face is often trying to be two menus. Count the dishes that will remain after the cuts. Under about 28 items, an A5 single sheet or a short A5 bifold usually reads better on a small table. Between 28 and 45, a single-sided A4 or a light A4 bifold holds the hierarchy. Above that, you are not redesigning a menu; you are refusing to edit. ISO 216 sizes (A4 210 × 297 mm, A5 148 × 210 mm) keep reprints cheap and holders standard. US Letter only if you already own Letter holders and will not change them.

Paper is a service decision. 300 to 350 g matte board with a soft-touch film survives a season of wet tables better than 170 g gloss that curls and fingerprints. If you laminate, accept that you cannot reprint in two weeks when the catch changes. If you reprint often, skip lamination and budget for a new run. A typical short EU run of 50 to 100 menus on 350 g board sits in the low hundreds of euros once design is settled; the expensive part is reprinting twice because the first brief was vague.

Write a one-page brief, then open the file

The brief should fit on one A4. If it does not, you are not ready. Include the surviving dish list in section order, the two or three dishes you want the eye to meet first, the dishes you will not highlight even if they sell, the format, the paper, the languages on the card, and three words for the mood. Not a paragraph of brand poetry. Three words a typesetter can use: warm, tight, no script. Or: bright, chalk, generous margins.

1. Surviving dishes in final section order, with prices and allergen marks already decided.
2. Two stars to place early in their section, and one puzzle you will give a short extra line.
3. Format (A4 or A5, single sheet or bifold, one side or two) and a paper weight.
4. What must stay from the old card: a logo lockup, a colour, a heading style the regulars would miss.
5. What must die: the eight-line history, the clip-art border, the 8 pt wine dump on the back.

Only then open a layout. If you generate with AI, paste the dishes in that order; the generator will follow the sequence you give it. If you work with a designer, the brief is the job. Either way, the first proof should be judged against the worn photographs and the till sheet, not against a mood board from another city.

Reprint when the decisions are boring

A good redesign feels slightly disappointing in the meeting, because the drama happened on the spreadsheet. The new card is shorter, the type is larger, the kitchen is not being asked to cook a lie, and the prices sit in a column that does not wobble. Print a paper proof at 100 %, fold it, sit at a table, and read it with the house lighting. If you still want to add four dishes to fill a hole, you have not finished the edit.

In Menu Atelier you paste the surviving list, drop a logo and a couple of room photos, and generate a layout you can turn in 3D before you spend on board. Switching from an overcrowded A4 single sheet to an A5 bifold takes one click once the list is honest; the print bundle ships with 3 mm bleed and crop marks, and if you order print from us the menus come on 350 g matte board with a soft-touch finish, scored on bifolds, in 5 to 8 working days across the EU.