Design and test restaurant menus

A menu should make choices clear before it tries to direct attention

A practical menu has a readable type size and contrast, clear sections, consistent item names, concise descriptions and a layout that works on paper and on a phone. Accurate prices, dietary information and allergen guidance support informed choice. The design should fit the restaurant and its customers, but visual style should not make important information hard to find.

Boxes, borders, colour, type changes and images can act as ‘eye magnets’ by drawing gaze toward an area. Pavesic (2005) discusses these as menu-design recommendations. They are not proof that a highlighted item will be ordered: attention, liking, choice, sales and profit are different outcomes.

Too many highlights, small print, clutter, misleading photographs or unclear categories can increase search effort and reduce legibility. A menu may be tested by comparing one design feature at a time. Eye-tracking can record fixations and scanpaths; researchers should also record item choice or orders and ask customers how clear the menu felt.

A fixation is evidence of where the eyes paused, not a direct measure of understanding, preference or purchase. Eye-tracking results depend on the task, menu format, hunger, familiarity and the participant’s goal. Pair gaze data with behaviour and self-report instead of treating it as a complete explanation.

Dayan and Bar-Hillel (2011) tested position in simultaneous printed menus

Primacy and recency usually refer to an advantage for items encountered first or last in a sequence. A printed menu presents category items together, so choosing a first or last item can also reflect spatial position or a preference for items near the edges. Dayan and Bar-Hillel tested this distinction rather than assuming the familiar serial-memory effect would apply.

In Study 1, 240 Hebrew University students were randomly assigned to one of four printed-menu orders. The menu offered the same foods and descriptions, with no prices; participants chose one item from each category. Items near the beginning or end of a category were chosen more often than the same items in the middle. The researchers reported an overall edge advantage, but no top-half versus bottom-half advantage: the result was not evidence of a simple primacy or recency effect.

Study 2 alternated a standard café menu and a version changing item positions, then recorded real orders in three categories. The overall edge-position advantage was similar, though category-level evidence was not equally strong for every category. The real café orders add everyday relevance, but customers, preferences and service conditions were not controlled like a laboratory task.

The studies concern the menus and restaurants tested, not every cuisine or digital interface. Study 1 used hypothetical selections without prices; Study 2 recorded actual orders but only at one café. A retailer should test its own menu and customers rather than assuming that putting an item at an edge will always increase sales.

Lockyer (2006) examined how words shape menu choice

Lockyer used four focus groups to compare five menus containing the same items but different wording, then used the focus-group themes to develop a broader survey. Participants discussed the dining occasion, companions and the appeal of dishes they might not make at home. Descriptive words such as ‘tender’, ‘golden’ and ‘natural’ were viewed positively in the survey.

The work provides both qualitative comments about why wording appeals and survey responses about preferred choices. It supports the idea that a menu description can shape expectations and stated selection. It does not establish that any particular adjective causes more purchases in every restaurant: language, food familiarity and occasion may differ, and a survey choice is not the same as a recorded order.

Descriptions should be vivid but accurate. A field test can compare two truthful descriptions of the same dish while keeping price and placement constant; record actual orders and ask a separate rating question. Do not use wording that misrepresents ingredients, origin, dietary suitability or portion size.

Match each method to the question

Eye-tracking is useful when the question concerns what a diner looks at and in what order. A questionnaire can ask what seems clear, appealing or confusing. A controlled menu task can isolate one design variable, while a field experiment can test real orders in a restaurant.

Objective outcomes include first fixation, time to locate an item, choice and recorded orders. Subjective outcomes include ratings and comments. Neither type alone answers every question: a diner can look at a dish without choosing it, and a stated preference may differ from a purchase.

A valid test changes one feature, uses comparable dishes and prices, includes a suitable comparison menu and records the same outcomes. Random assignment strengthens a causal claim; an uncontrolled before-and-after sales comparison may be affected by day, season, stock, price or customer mix. Results from one sample or restaurant should be generalised cautiously.