For most of restaurant history, the only data a menu produced was the bill. You knew what sold because you counted it at the end of service. You never knew what guests considered and rejected, what they read twice and put down, or which corner of the menu they never reached at all. The menu was a one-way object: you printed it, the guest read it, and everything that happened in between stayed invisible.
A digital menu changes that quietly. Because guests browse it on their phones, it can record what they look at before they decide, not just what they end up ordering. That stream of behaviour is restaurant menu analytics, and used well it is one of the most underrated tools a restaurateur has. Used badly, it is just another dashboard nobody opens.
This guide is about using it well. We will cover the handful of numbers that actually matter, what each one is really telling you, and the specific change to make when a dish underperforms. No vanity metrics, no charts for the sake of charts.
What menu analytics actually measures
Start with the distinction that makes the whole thing useful: sales data versus behaviour data.
Sales data is what the till already gives you. How many of each dish left the kitchen, at what price, on what day. Every restaurant has this, paper or digital. It tells you the outcome.
Behaviour data is what a digital menu adds. Which categories guests opened, which items they tapped to view, how long they spent, which language they switched to, what device they used, and crucially, which dishes attracted attention but were never ordered. It tells you the journey that led to the outcome.
The combination is where the value is. Sales data alone tells you the seabass sold twelve times last week. Behaviour data tells you the seabass was viewed ninety times and ordered twelve, while the lamb was viewed thirty times and ordered eleven. Same rough sales, completely different stories: the lamb converts beautifully and just needs more visibility, while the seabass pulls attention and loses it at the final step. Those two dishes need opposite interventions, and only the behaviour layer reveals which is which.
The five numbers worth watching
You could track dozens of metrics. You should track five. Everything else is context for these.
1. Item views
The number of times guests opened a specific dish. This is raw demand and curiosity. It tells you what draws the eye, which photos pull people in, and which dishes are effectively invisible. An item with almost no views is not failing to sell because guests rejected it; it is failing because guests never saw it. That is a placement and presentation problem, not a recipe problem, and the fix is far easier.
2. View-to-order ratio (conversion)
Of the guests who viewed a dish, how many ordered it. This is the single most diagnostic number on the menu. A high view count with a low conversion ratio is the classic signal that something is wrong at the moment of decision: price, description, photo, or missing information. A modest view count with a high conversion ratio is the opposite, a quiet winner that deserves a better spot on the menu.
3. Item profitability
Not price, margin. The dish that earns you the most per plate is rarely the most expensive one; it is the one with the best gap between menu price and food cost. Analytics that show you popularity without profitability lead to the trap of promoting your best-selling loss leader. Cross popularity with margin and every dish falls into one of four boxes, which is the heart of menu engineering and the subject of our menu design tips.
4. Category drop-off
How far down the menu guests actually travel. Most menus are read top to bottom and attention decays as guests scroll. If desserts get a fraction of the views starters get, that is not necessarily because nobody wants dessert; it is often because the dessert section sits at the bottom of a long page and most guests have decided before they reach it. Drop-off tells you where the menu loses its audience.
5. Language and device mix
What share of guests switch to another language, and what they read on. This is the cheapest insight to act on and the most often ignored. If a third of your guests view the menu in another language, your tourist business is bigger than your bookings suggest, and your descriptions need to translate as well as they read. Our guide to multilingual menus covers how to handle that without machine translation backfiring.
How to read the patterns
Numbers on their own change nothing. The skill is matching a pattern to a decision. Here are the four you will see most often and what each one is asking you to do.
High views, low orders
The dish attracts attention but loses the guest at the last step. Work through the causes in order of likelihood. First the photo: a weak or missing image is the most common silent killer, and a strong one is the cheapest fix. Our smartphone food photography guide covers shooting an entire menu in one morning. Second the description: vague wording ("served with seasonal vegetables") converts worse than specific, sensory wording. Third the price: if photo and description are strong and it still stalls, the price may be sitting above what the guest expects for what is described. Change one thing at a time so you know which lever moved the number.
Low views, high orders
A hidden gem. Guests who find it love it, but too few find it. This is the easiest win on the whole menu: move it up, give it a photo if it lacks one, or flag it as a recommendation. You are not changing the dish, only its visibility, and the conversion is already proven.
High views, high orders, low margin
Your popular workhorse, and a quiet drag on profit. You do not remove it, guests love it, but you can engineer around it: a well-priced add-on, a pairing suggestion, or a slightly higher-margin variant placed right beside it. This is where analytics meets upselling done as hospitality, nudging the average ticket up without pressure.
Low views, low orders, every week
A dead weight. A dish that nobody looks at and nobody orders, week after week, is costing you in prep, stock, and menu clutter. Every item you cut makes the rest of the menu easier to read and faster to decide from. Analytics give you the confidence to remove a chef favourite that the room has quietly voted off.
A simple weekly and monthly rhythm
Data only helps if you look at it on a schedule. The trap is either never opening the dashboard or opening it obsessively and reacting to noise. A light, regular rhythm beats both.
| Cadence | What to look at | What to decide |
|---|---|---|
| Weekly (15 min) | Top and bottom five by views; anything that suddenly dropped | Quick fixes: fix a broken photo, promote a hidden gem, pull a sold-out item |
| Monthly (45 min) | View-to-order ratios, margin per dish, category drop-off, language mix | Real menu changes: reword, reprice, reorder, or remove dishes |
| Quarterly | Seasonal shifts, full menu engineering pass | Refresh the menu structure, plan the next season's dishes |
One discipline matters above all: give every change time. Menu data only stabilises across a few hundred covers, so a single slow Tuesday means nothing. When you reword a dish or move it up the page, leave it for two to four weeks before you judge whether the change worked. Reacting day to day turns analytics into anxiety and teaches you nothing.
The analytics most owners never think to use
Two patterns sit outside the standard list but repay attention.
Time-of-day behaviour. The same menu is read very differently at lunch and at dinner. Lunch guests scan fast and convert on familiar dishes; dinner guests browse longer and explore more. If your data shows lunch viewers barely reaching the mains, a tighter lunch view, or a lunch-specific section pinned to the top, can lift the midday ticket without touching the evening menu.
Search and dead ends. If your menu has a search box, what guests type into it is a gift: it is the list of things they expected to find and did not. Repeated searches for "vegan," "gluten free," or a specific dish tell you exactly what is missing or badly labelled. The same logic applies to allergen filters, which double as a record of what your guests actually need to avoid.
From data to a found, machine-readable menu
There is a second payoff to a structured, digital menu that has nothing to do with the in-room dashboard. A menu that is properly structured is also readable by search engines and AI assistants. When someone asks Google, ChatGPT, or Claude about your restaurant, a machine-readable menu with prices, allergens, and descriptions is the most complete answer on the internet, better than a photographed PDF or a third-party listing. The same data that helps you tune the room helps guests find you in the first place.
This is where having the analytics and the public menu in one place matters. At Opnclo, the menu guests scan, the analytics you act on, and the structured page that Google and AI assistants read are the same object. You are not bolting a tracking tool onto a static menu; the data is a by-product of running the menu, and the menu is already built to be found. If you are still weighing the format question, our digital versus paper comparison lays out the trade-offs, and the complete guide to QR code menus covers everything restaurateurs ask before switching.
Start small, act on one thing
The most common mistake with menu analytics is treating it as a research project. You do not need a data team or a wall of charts. You need to look at one screen once a week, find the single clearest signal, and make one change. Next month, look again and see if the number moved.
Pick your worst high-views-low-orders dish and fix its photo. Find your best hidden gem and move it up. Cut the one item nobody has viewed in a month. Three small, evidence-based decisions a month compound into a menu that quietly works harder than it did, season after season, without you guessing.
Frequently asked questions
What is restaurant menu analytics?
It is measuring how guests interact with your menu and using that to make decisions. A paper menu gives you only sales totals from the till. A digital menu also shows what guests looked at before ordering: which categories opened, which dishes were viewed, how long people stayed, and which items drew interest but few orders. Combining view data with sales data tells you not just what sold, but why.
Which menu metrics actually matter?
Five carry almost all the weight: item views, view-to-order ratio, item profitability (margin, not price), category drop-off, and language and device mix. Most other numbers are interesting but rarely change a decision.
What does it mean when a dish gets many views but few orders?
The dish attracts attention but something stops the guest at the last step, usually the photo, the description, or the price. Test one change at a time, starting with the photo and description, before assuming the price is the problem.
How often should I review my menu analytics?
A fifteen-minute look once a week for anything obvious, and a deeper review once a month for real menu changes. Do not react to a single slow day; give any change two to four weeks before judging it.
Do I need a separate analytics tool?
No. If your menu is digital, the analytics should be built in. A QR-code menu platform records views, orders, language, and device automatically, so the data comes free with running the menu. A paper menu gives you none of this, which is the main reason data-minded owners move to digital.