# Menu naming checklist

The food does not change. Only the words do. These are the rules the evidence supports, with the figures behind them, so you can argue for them with whoever signs off the menu.

*The Balance of Health. Generated 2026-09-06. Sample data pending verification.*

## The rules

- Name the dish, never the category. 'Smoky black bean burger', not 'veggie burger', and never 'meat-free burger'.
- Lead with flavour, texture or method: smoky, slow-roast, charred, crisp, sticky, buttered.
- Do not name what is absent. Describing a dish by what it lacks invites the reader to feel the loss.
- Do not name it for health. The same drink was rated markedly less enjoyable when labelled healthy.
- Name the origin or the tradition where it is real: chana masala, minestrone, dal. These already sound like food.
- Keep the promise the name makes. A name that oversells costs more in disappointment than it gains at the counter.
- Put plant dishes in the same part of the menu as everything else, not in a separate section.

## The evidence

| Finding | Figure | Source |
| --- | --- | --- |
| Rise in vegetarian meal sales when vegetarian options doubled, observational | 15 percentage points | University of Cambridge (2019) |
| Rise in vegetarian meal sales when vegetarian options doubled, experimental | 7.8 percentage points | University of Cambridge (2019) |
| Odds of a vegetarian choice when availability doubled | 2.6 odds ratio | French university catering research group (2023) |
| Diners who noticed the menu had changed | 6 % | French university catering research group (2023) |
| More often chosen under a flavour label than a health-restrictive one | 41 % | Stanford University (2019) |
| More often chosen under a flavour label than a basic one | 25 % | Stanford University (2019) |

## What this will not do

- It will not work on every dish. Effect sizes vary a lot between dishes and settings.
- Menu-sequence and dish-of-the-day nudges were tested in a similar setting and changed nothing.
- Selection at the counter is what was measured, not what was eaten or what happened next time.