Menu Analytics: What Guests Look At but Never Order (2026)

Your sales report tells you what worked. It never tells you what almost worked. Somewhere on your menu there's a dish guests open, read, think about — and then quietly skip. Every single night.

Here's the short version. Menu analytics is the guest side of your data: it tracks what people scroll past, tap, open and abandon on your digital menu, not just what lands on the bill. The number that matters is the gap between views and orders. A dish with heavy views and almost no sales isn't a kitchen problem. It's a price, wording, photo or position problem — and all four are free to fix. You find the gap, change one thing, and test it for 14 days.

Most owners have never seen this data, because a paper menu can't produce it. A digital menu can, and in 2026 that's the real reason to run one.

What Is Menu Analytics?

Menu analytics measures how guests behave before they order.

Your POS reports are a record of decisions already made. Menu analytics is a record of decisions being made — the browsing, the comparing, the second thoughts. Put simply:

	Sales reports	Menu analytics
Measures	What guests bought	What guests looked at
Source	Orders and payments	Digital menu views and taps
Answers	"What sells?"	"What nearly sold?"
Blind spot	Interest that never converted	Anything ordered out loud

You need both. Sales data alone will tell you a dish is a loser and you'll delete it — when the truth might be that 300 people opened it last month and the price scared off 297 of them.

That's not a dead dish. That's a dish with a broken offer.

The Guest Menu Funnel: Four Steps, Three Places to Lose Them

A guest on a digital menu moves through a small funnel. Most tools track it as four events:

1. Scrolled to a category — the menu loaded and they went past this section
2. Selected a category — they chose to look inside
3. Viewed product details — they opened one specific dish
4. Added to cart — they committed

Between each step, people fall away. That's normal. What's useful is where they fall away, because each gap has a different cause.

• Lost between scroll and category? Your category names are boring, unclear, or in the wrong order.
• Lost between category and product? The dish name and price in the list aren't pulling anyone in.
• Lost between product view and cart? This is the big one. They read the whole thing, and something in it stopped them.

That last drop-off is what this article is about. Real interest, no sale.

The Four Numbers to Pull First

You don't need a data analyst. You need four numbers and about twenty minutes a month.

1. View-to-order rate (the one that matters)

View-to-order rate = Orders of a dish ÷ Product detail views of that dish

Some people call this the look-to-book rate, borrowed from hotels. Call it whatever sticks with your team.

A worked example. Last month your lamb shank got 412 detail views and sold 18 times. That's a 4.4% view-to-order rate. Your burger got 380 views and sold 141 times — 37%.

Same menu. Same guests. One of those dishes is losing people at the last second.

Ignore anyone quoting you an industry benchmark for this. It swings wildly by venue type, price point and how the menu is built. Your own menu average is the only benchmark worth having. Work it out, then hunt for the items sitting well below it.

2. Category reach

What share of visitors ever open each category? If your desserts are opened by 9% of guests, you don't have a dessert problem. You have a placement problem — and probably an average ticket problem attached to it.

3. Cart abandonment

Items added and then removed, or carts started and never sent. On a self-ordering or delivery menu, this is often a delivery fee, a minimum order, or a total that suddenly looks bigger than expected.

4. Traffic context: device, country, source

Where did the scan come from, and who's holding the phone? A menu with 30% of traffic from outside your country needs multilingual QR menus far more than it needs a new dessert. Tourists don't ask what a dish is. They just order the thing they recognise.

Why Guests Look at a Dish and Never Order It

After enough menus, the same eight causes come up again and again. Roughly in order of how often they're the real culprit.

1. Price shock at the wrong moment

On a phone, the price often lands after the description — right at the end, when the guest has already imagined eating it. That mismatch between what they expected and what they read is what kills the order.

Sometimes the price is genuinely too high for the slot. More often the dish just isn't carrying its own justification. A $34 short rib with a nine-word description reads as expensive. The same dish with "slow-braised for 8 hours, feeds two comfortably" reads as good value.

Before you cut a single price, be sure the number is actually the problem — pricing and perceived value is the deeper version of this argument.

2. The description doesn't answer the three silent questions

Every guest is quietly asking:

• How big is it? Will I still be hungry?
• What comes with it? Do I have to buy a side too?
• What is this, exactly? Spicy? Raw? Sweet? Weird?

If your description doesn't answer all three, some people close it and pick something safer. They will never tell you this. The analytics will.

3. No photo — or a photo doing damage

On a digital menu, an item without a photo sits next to items with photos and loses. That's not a taste judgment, it's just how a scrolling screen works.

But a bad photo is worse than none. Dark, yellow, shot from above on a dirty pass at 11pm — that image is actively costing you orders. Worth reading how to shoot menu photos for online ordering before you photograph anything, because half-doing it makes things worse.

4. Allergen and diet uncertainty

A guest with a nut allergy, a coeliac diagnosis or a vegetarian partner does not gamble. If the dish doesn't say, they move on — silently, and often for the whole table.

This one shows up beautifully in analytics: high views, near-zero orders, on dishes where the ingredients are ambiguous. Fix it by declaring allergens on the digital menu properly, per dish, in every language you publish.

5. It's buried

Position beats quality on a phone screen. An item at the bottom of a 40-item category gets a fraction of the views of the top three — and the ones who scroll that far are usually looking for something specific.

If a dish has low views and low sales, it isn't unpopular. It's invisible. Completely different problem, completely different fix.

6. Choice overload

Fourteen pasta dishes don't sell more pasta than six do. They sell less, because a tired guest facing too many options defaults to the safest, cheapest, most familiar thing on the list.

You'll see this as flat, low view-to-order rates across a whole category, not one bad item.

7. It's 86'd — and still on the menu

Guests choose a dish, get told no, and re-order defensively. The second choice is almost always cheaper. If your menu can't be switched to "unavailable" in ten seconds, this happens more than you think.

8. It's in a language they can't read

Different from tourists — this is the guest who can read the words but doesn't know the dish. "Sobrasada", "nduja", "sumac". Great words. Add four more explaining them.

The Opposite Signal: Ordered a Lot, Barely Looked At

There's a second pattern worth hunting for, and it's easy to miss.

Some dishes sell well with very few detail views. Nobody opens them, but they get ordered constantly. These are habit items — the flat white, the house burger, the kids' pasta. Guests already know what they want and never open the page.

Two rules for these:

• Don't "improve" them. Nobody is reading your new description. They're tapping from memory.
• Don't move them. A habit item that gets relocated becomes a lost item, and it takes weeks for anyone to tell you.

The place to invest effort is the item with lots of looking and little buying. That's where interest already exists and something small is blocking it.

What to Do About It: The Fix List

Ranked by effort, cheapest first. Do them in this order.

1. Rewrite the description. Free, takes ten minutes, and it's the most common fix that works. Answer the three silent questions. Name the origin, the method, or the time it takes. Use the words guests would use, not the words a chef would use.

2. Move it up. Top three positions in a category, or into a category people actually open. Costs nothing on a digital menu. This alone can double views.

3. Add or replace the photo. One good daylight shot beats ten bad ones. Photograph your top five and your problem items first — the rest can wait.

4. Add the missing information. Portion size, allergens, spice level, what's included. Every unanswered question is an exit.

5. Give the price a companion. People judge price by comparison, not in isolation. A $34 dish next to a $41 dish reads as reasonable. Alone at the top of a page, it reads as the expensive one.

6. Change the price — last, and carefully. Small moves. And check your plate cost before you touch anything, or you'll cut into margin you can't afford; how to calculate plate cost covers the maths.

7. Bundle it. A slow-moving dish with a good margin often flies as part of a set — dish, side and a glass of wine at a single price.

8. Delete it. If the views are strong and nothing you try converts, the dish is telling you something. Free up the space, free up the prep, keep the recipe for a special.

Change one thing at a time. Two changes at once and you'll never know which one worked.

The 14-Day Menu Analytics Test

This is the whole method, and it takes about half an hour of actual work.

Day 0 — Get a baseline. Pull last month's views and orders for every dish. Calculate view-to-order rate for each. Write down your menu average. Screenshot it, or export the report to PDF so it can't be argued with later.

Day 0 — Pick your three worst offenders. Not the three lowest sellers. The three with the widest gap between views and orders. High interest, no sale.

Day 1 — Change one thing on each. One dish gets a new description. One gets moved up. One gets a photo. Note the date, in writing.

Days 2–13 — Leave it alone. Genuinely. The urge to keep tweaking is what ruins these tests.

Day 14 — Compare. Same day-count, same days of the week. Did the view-to-order rate move? A jump from 4% to 9% on a $29 dish with 400 views a month is roughly 20 extra covers — around $580 in sales for zero cost.

Then repeat with the next three. Six weeks of this and you'll have gone through your whole menu once.

One warning on sample size. If a dish only got 22 views last month, the numbers mean nothing. Wait until you have a few hundred views before you trust a rate. Small numbers move dramatically for no reason at all.

Menu Analytics Plus Menu Engineering Equals the Full Picture

Classic menu engineering sorts dishes by profit and popularity into Stars, Plowhorses, Puzzles and Dogs. It's a great model. It has one blind spot: it only sees dishes people bought.

Add view data and the Puzzles split into two completely different animals:

Views	Orders	What it really is	What to do
High	High	A genuine star	Protect it, never touch the recipe
High	Low	A broken offer	Rewrite, re-shoot, reprice — the fix is free
Low	High	A habit item	Leave it exactly where it is
Low	Low	Invisible, not unwanted	Move it up first, then judge it

That second row is the one nobody sees without menu analytics. It's usually where the fastest money in the whole exercise is sitting.

And the two systems feed each other. Menu analytics tells you why a Puzzle is a Puzzle. Menu engineering tells you whether it's worth saving.

What Menu Analytics Can't Tell You

Be honest about the limits, or you'll over-read the data.

• It doesn't see the table conversation. "Don't get that, it was dry last time" is invisible and powerful.
• It doesn't see your waiters. A recommended dish sells for reasons that have nothing to do with the menu. Your floor team is still the strongest lever you own — upselling technique moves more revenue than any layout change.
• It doesn't cover paper. If half your guests use a printed menu, you're reading half a story.
• It can't tell curiosity from intent. Some people open the $95 tasting menu purely to look. That's fine — just don't expect it to convert.
• It's useless at low volume. A café doing 40 covers a day needs a full month before any rate is meaningful.

Use it as a shortlist generator, not a verdict. The data says "look here". You still have to decide what's wrong.

Guest Privacy: What You Should Actually Be Collecting

Good menu analytics is aggregate and anonymous. Counts of views, taps and scrolls. No names, no phone numbers, no tracking anyone across the internet.

You do not need personal data to learn that your lamb shank isn't converting. If a tool is offering to identify individual diners, that's a very different product with very different obligations.

Rules on cookies and consent differ by country and keep changing — in the EU and UK especially. Check your setup against your own regulator's current guidance, or ask a local advisor. A short conversation now is cheaper than a complaint later.

Menu Analytics FAQ

What is menu analytics? It's the measurement of how guests interact with a digital menu — which categories they open, which dishes they view, and which ones they add to an order. It shows interest that never became a sale, which sales reports can't do.

What's a good view-to-order rate? There's no universal number, and anyone quoting one is guessing. Calculate your own menu-wide average, then treat anything far below it as a candidate for a fix.

Why do guests look at a dish but never order it? Usually price shock, a description that leaves questions unanswered, a missing photo, unclear allergens, or a bad position on the page. In most cases it's the offer, not the food.

Can I track a printed menu? Not really. You can infer things from sales mix, but views, scrolls and abandoned choices only exist on a digital menu. It's the single biggest data advantage of a QR menu over paper.

Do I need a POS to use menu analytics? No. Menu analytics comes from the guest-facing menu. A POS adds the other half — what actually sold — and the two together are much stronger than either alone.

How much traffic do I need before the numbers mean anything? A few hundred views per dish is a reasonable floor. Below that, one busy Saturday can flip the ranking completely.

Does menu analytics cost money? It shouldn't. Plenty of digital menu tools include it. For what it's worth, Tabres tracks the four guest events above across eight date ranges, plus sources, countries and devices — alongside best-seller, worst-seller and trending reports and add-on attach rates. It's free, with no per-branch fee, and our reasoning for that is public. Use whatever tool you like; the method in this article works anywhere.

My menu is a PDF. Can I get this data? Not from a PDF — a file can't report anything back. You'd need to convert it into a real digital menu first, which is a one-afternoon job these days: turning a PDF menu into a QR menu walks through it.

---

Most restaurants are still making menu decisions with half the evidence. They see the sales, delete the slow movers, and never find out that the "failed" dish was opened four hundred times by people who wanted it and then changed their mind.

That gap between looking and ordering is the cheapest money left in hospitality. No new guests, no ad spend, no new equipment — just better answers to questions your guests are already asking silently, on their phones, at your tables.

Pull one report this week. Find the dish with the widest gap. Change one thing about it, and check back in fourteen days.