A menu builder your kitchen can change on a Tuesday afternoon

Menu structure advice: the mistakes that cost orders

The menu builder handles the mechanics. The decisions below are the ones that decide whether the page sells. None of them need software; they need an hour with the menu and a critical eye.

Too many items

The dining-room menu exists partly to look generous. The online menu exists to be ordered from on a phone in under two minutes. A menu of ninety items forces scrolling, hides what sells, and means more things to keep in stock and keep available. Start with the forty that travel well and sell most. Add back by exception.

Categories that match the kitchen, not the customer

“Chef’s specialities” means nothing to someone choosing dinner. Name categories the way a customer would ask for things at the counter: starters, biryani, curries, breads, rice, desserts, drinks. Put the highest-margin category where the eye lands second, after the category people search by name.

Sizes hidden in the description

“Available in half and full” in the description, with one price, is the single most common cause of wrong orders. If sizes exist, they are variants with their own prices. The customer picks one, the price builds correctly, and the kitchen ticket says which.

Add-ons that are really required choices

If every curry needs a spice level, set it up as a variant of the dish at the same price (mild, medium, hot) rather than a note saying “tell us how spicy”. If a thali must come with a bread choice, the same applies. Optional add-ons are for things that cost extra.

Descriptions written for people who already know

A regular knows what gobi 65 is. A new customer does not and will not look it up. One line answers the question: crisp cauliflower florets in a spiced batter, medium-hot, serves one as a starter. It takes thirty seconds to write and it sells the dish for years.

Prices that do not match the receipt

Every price on the menu is the price at checkout. If a delivery charge applies, it is shown before payment, on its own line, not folded into item prices only on delivery orders. The moment a customer sees a different total than they expected is the moment they close the page.

Photos that hurt

A dark, blurry photo lowers the perceived value of the dish below what no photo would. If you have a good photo, use it; if not, leave the item without one and let the description do the work. When you do photograph, do the dishes you most want to sell first, in daylight, from slightly above.

Nothing scheduled, everything manual

If breakfast ends at 11 and someone has to remember to hide it every day, it will be visible at 11:40 on the day it matters. Schedules exist so the menu runs itself. Set them for breakfast, for weekend-only items, and for specials with an end date.

Light menu engineering

You do not need a spreadsheet to do the basics. Once a month, look at which items sold most and which sold least. Put the popular, high-margin items near the top of their category with a photo. Consider dropping the items that sold under a handful of times unless they matter to regulars. Raise the price of the item everyone orders anyway by a small amount before raising anything else. The guide to encouraging repeat orders covers how that interacts with the direct channel.