One Dish, Four Prices: Menu Price Divergence

You set one price for the city. The customer opens two delivery apps and sees two numbers. In August 2026, the price of the same dish from the same restaurant matched across UAE platforms only 72–85% of the time, depending on the pair, and the gaps were not random: in three of the four markets one platform sat consistently on one side. “One price” is a policy inside your company. On the storefront it is a fact you have to check.

Your one price is not what customers see

Your price list says one thing. Your storefronts say several. Among restaurants monitored by Kitchain in the UAE, we took every dish of every restaurant that was listed on two or more platforms and compared the price shown on each app on four August days. Here is how often the price of the same dish matched, pair by pair:

• Careem and Talabat: matched 85% of the time.
• Careem and Deliveroo: matched 73% of the time.
• Noon and Talabat: matched 72% of the time.

A match means the identical price on both apps. Anything else counts as a difference. So even on the best-aligned pair in the UAE sample, a noticeable share of dishes shows the customer a different number depending on which app they open, and on the least-aligned pair the share is far larger.

The point for an owner is simple: the price list is a document, the storefront is what the customer sees, and in August the two agreed less often than a single price list implies.

What this means for a restaurant: if you have never opened your own listing on two apps side by side and compared the same dish, you do not know your price. You know your policy.

The gap has a direction

The differences are not noise scattered evenly across platforms. In the UAE, the UK and Kuwait they lean one way, and once you know the lean you know where to look first.

In the UAE, where the price of a dish differed between two platforms, Talabat was usually the cheaper of the pair. At the other end, in 70% of dish groups the price on Careem was at least as high as on any other platform in the group. Noon was second, at 18% of groups. Those are shares of groups across the sample, not a verdict on any platform, and they say nothing about why the prices differ. They do say that a customer comparing apps in the UAE will most often find your dish at its highest price on Careem and at a lower price on Talabat.

For a chain, direction is the useful part. A random scatter would mean checking everything. A lean means the first check is the pair most likely to disagree, and the first question is whether your own listings follow the market’s pattern or break it.

What this means for a restaurant: “Careem is expensive” is not a finding you can act on. “In 70% of groups the price on Careem was the highest in the group, and here is the list of our own dishes where that holds” is.

Many gaps, all shallow: Saudi Arabia

Saudi Arabia looks like the opposite problem on the surface: prices differ often, but by very little. The risk sits in the tail, not the median.

Between HungerStation and Keeta, the price of the same dish differed for 57% of dishes in three August snapshots. But the gaps were small: the median spread was 3.3%, and nine in ten gaps were under 27%. There was no consistent cheaper side either: HungerStation was the cheaper of the two 47% of the time, close to an even split. Across the Saudi market as a whole, 44% of dishes showed no spread at all and another 27% showed a spread under 10%.

Two readings follow, and both are true:

• The single-price policy largely holds. 44% of dishes sit at one price across platforms and another 27% within 10% of it, and the typical difference among the rest is a few percent.
• The exceptions are invisible in the average. “Nine in ten gaps under 27%” also means the rest are above it, and because neither platform is reliably the cheaper one, an owner cannot guess which app carries the outlier for a given dish.

What this means for a restaurant: in Saudi Arabia, do not report the median gap and stop. The median is reassurance. The work is finding the dishes with a gap above 27%, and that has to be done dish by dish and branch by branch, because the direction flips.

One platform apart: the UK and Kuwait

In the UK and Kuwait the pattern is the same shape: two apps carry your price list, and the third carries a different one for more than half the menu.

In the UK, in two August snapshots, Just Eat and Uber Eats were price for price: the price of the same dish differed on only 3% of dishes. Deliveroo sat apart. It was cheaper than both for 56–61% of dishes, typically by 25%, meaning the median spread among the dishes that differed was a quarter of the lower price.

In Kuwait, in four August snapshots, Jahez and Talabat differed on 13% of dishes. Snoonu sat apart, and in the other direction: Jahez was cheaper than Snoonu in 72% of groups, typically by 57%, with the spread measured against the lower price.

The structure matters more than the platform names. In both markets, a chain that checks two of its three storefronts and finds them in agreement has learned nothing about the third, and the third is where the customer sees a different price.

What this means for a restaurant: in a two-agree-one-apart market, the one apart is the whole audit. Which platform it is, and in which direction, is a per-market fact to look up, not to assume from another country.

What this means for a single-price chain

Three moves follow from the data, each with a scenario. Every scenario is a picture, not a client.

1. Treat the price list as a policy and the storefront as the fact. Check the storefront. Picture an 8-branch shawarma chain in Dubai with one price list, listed on Careem, Talabat, Noon and Deliveroo. Its 60-item menu is the same on paper everywhere. As a scenario, at the August match rate between Careem and Talabat of 85%, about 15% of dishes show a different price on the two apps, which on a 60-item menu is about 9 dishes (60 × 15% ≈ 9). Between Noon and Talabat, at a 72% match rate, it is roughly 17 dishes (60 × 28% ≈ 17). A customer who checks two apps before ordering sees the chain break its own rule on nearly a third of the menu. The owner, looking at the price list, sees one price.

2. Look for the direction in your market, then check whether your own listings follow it. Picture a three-branch burger brand in Kuwait City on Jahez, Talabat and Snoonu with one city price. As a scenario: Jahez and Talabat differ on 13% of dishes, so they agree on the other 87%. On Snoonu, in 72% of groups the same burger shows a price typically 57% above Jahez, which is more than one and a half times the price the owner set (1 + 57% ≈ 1.6×). Anyone comparing apps sees the brand as two different price points. The branch manager, looking at the POS, sees one. Knowing that Snoonu is the app that sits apart in this market turns a full audit into a targeted one.

3. Hunt the tail: the dishes with the biggest gap, branch by branch. Picture a grill chain in Riyadh on HungerStation and Keeta whose owner is relieved by a 3.3% median gap. As a scenario: 57% of dishes differ between the two apps, which on a 60-item menu is about 34 dishes (60 × 57% ≈ 34); nine in ten of those gaps are under 27%, so about 3 dishes (34 × 10% ≈ 3) sit above it, where the customer sees a materially different price. With no consistent cheaper side, since HungerStation was cheaper only 47% of the time, the owner cannot guess which app carries the outlier. The check has to be per dish, per branch, and it has to be repeated.

Kitchain Menu does this for you. It opens the same public storefront pages a customer opens, reads list and effective prices for every item in every branch on every platform, and shows the price parity gaps between branches and between platforms, so the outliers are on a list instead of in a customer’s screenshot. No integration, no platform login.

How we measured

The unit is a group: one dish of one restaurant, listed on two or more platforms. For each group we compared the price shown on each platform on a sample day. Our data for this issue does not distinguish list price from the price after any offer, so throughout this article “price” means the menu price shown on the app that day.

Definitions used above:

• Match. The identical price on two platforms. Any difference, however small, is a non-match.
• Match rate. The share of groups in a platform pair where the price matched. “Matched 85% of the time” means 85% of groups on that pair.
• Spread. (Maximum price − minimum price) / minimum price within a group. A spread of 57% means the higher price is 57% above the lower one.
• Median spread, or “typically”. The middle spread among the groups that differ. Groups that match are not in that median.
• “Cheaper in 72% of groups.” A count of groups, not of dishes across a menu, and not a share of revenue.
• “Nine in ten gaps under 27%.” A share of the gaps among dishes that differ, not of all dishes.

Sample and period:

• Restaurants monitored by Kitchain, with demo accounts that have real listings included. Russia is excluded.
• Sample days in August 2026: four in the UAE and Kuwait (7, 14, 21 and 28 August), three in Saudi Arabia (14, 21 and 28 August), two in the UK (21 and 28 August, as collection there started mid-month). UK figures are therefore “in two August snapshots”.
• Comparability. Groups that could not be matched across platforms were 0.3% in the UAE; no groups were quarantined.
• Two platform pairs are left out of this issue while we verify how their prices are collected.
• No month-on-month comparison. Menu prices have been collected since 6 July 2026, and the collection method changed on 1 September, so September cannot be compared with August. The figures here are an August snapshot.
• No causes. The data shows that prices differ, where, and by how much. It does not show why. Nothing in this article should be read as an explanation of the gaps.

Glossary:

• Group. One dish of one restaurant on two or more platforms.
• Spread. The gap between the highest and lowest price in a group, as a share of the lowest.
• Match rate. The share of groups on a platform pair where the price was identical.
• Median spread. The middle value of the spread among groups that differ; the “typical” gap.