Get the App Get the App
MenuClose
Get the App

The Delivery Markup You Can’t See — and Why It Breaks the Split

Jun 21, 2026 · 10 min read

Four friends. One $74 delivery order. Split four ways. But the prices you’re dividing were quietly marked up — 3% on one restaurant, 30% on another — and the receipt never says by how much.

The receipt you split isn’t the receipt the restaurant printed

Here’s the thing nobody mentions when four of you split a delivery order: the prices you’re dividing up aren’t real. Not in the way a dine-in check is real. A paper restaurant check lists what the kitchen charges. A delivery receipt lists what the restaurant decided to charge on the app — a number it quietly rewrote upward to survive the platform’s commission. Same burrito, different price, and the app never tells you the difference.

That gap has a name in the trade — menu markup — and on its own it’s well documented. We’ve already mapped where every delivery fee goes and why a $16 dish checks out at $30. This isn’t that. This is about what the markup does to the split. Because the markup is invisible, and because it isn’t the same size at every restaurant — or even on every item — dividing a delivery receipt fairly is harder than it looks. You’re splitting a number that has already been bent, by an amount no one at the table can see.

3%–30%The range of delivery menu markups across major chains — a casual-dining spot might add 3%, a high-demand brand nearly 30%. Same channel, ten-fold difference, none of it labeled.

Why the menu price changes when you order delivery

Start with the engine. When a restaurant lists on a delivery platform, it pays the platform a commission — “typically a percentage of the order amount,” as the operations researchers Jaelynn Oh, Chloe Glaeser, and Xuanming Su put it in a 2023 Management Science study of how the two sides divide the money. That commission is steep, and it just got steeper: on March 10, 2026, Uber Eats raised its U.S. marketplace rates, pushing its Lite tier from 15% to 20% and charging 30% on orders from Uber One members.

A restaurant running on thin margins can’t absorb a 20–30% cut on every order. So it does the obvious thing: it raises the menu price on the app to claw the commission back. The customer pays the markup; the restaurant nets roughly what it would have in store. Nothing about this is secret in principle — it’s just never shown to you as a line.

Why this matters for your split: The markup isn’t a fee the platform tacks on at the end, like delivery or service charges. It’s baked into the price of the item itself. That’s what makes it invisible — and what makes it land unevenly, because every restaurant sets it differently.

And here’s the detail that turns one invisible markup into a fairness problem: platforms don’t charge every restaurant the same commission. They offer “the choice of a range of commission rates,” the Management Science authors note, “rewarding higher commission payments with featured display slots and discounted delivery fees.” Restaurants self-select into different cost structures. Different commissions, different margins, different markups. There is no single “delivery tax.” There’s a thousand of them.

Sources: Oh, Glaeser & Su, “Food Ordering and Delivery: How Platforms and Restaurants Should Split the Pie,” Management Science (2023); Restaurant Dive, “Uber Eats raises marketplace fees” (March 2026).

The markup isn’t one number — it’s a different number everywhere

This is the part that breaks the split. When Gordon Haskett Research Advisors tracked delivery menu prices across 25 chains, it didn’t find a tidy, uniform premium. It found a spread. Fast-food menu prices ran more than 15% higher for delivery than for pickup on average — but that average hides enormous variation by restaurant.

Restaurant typeExampleDelivery markup vs. in-store
High-demand QSRChick-fil-A~30%
National QSRStarbucks, McDonald’s~20%
Fast casualPanera-type~12%
Casual diningSit-down chains~3%

Chick-fil-A marks delivery prices up nearly ten times as much as a casual-dining chain does. Why the spread? Pricing power, mostly — a brand people will chase can pass more of the commission on to the customer. But margins matter too. Gordon Haskett’s analysts noted that Chipotle would clear just $1.10 on an order of 20 burritos at a 15% commission, versus about $4.10 on the same order for pickup. A restaurant that thin has to mark up aggressively or lose money on every delivery. A restaurant with fatter margins can leave the price almost alone.

The takeaway for anyone splitting: the markup buried in your delivery receipt could be 3% or 30%, and the app gives you no way to tell which. “Split the receipt fairly” quietly assumes the prices on it mean the same thing. Across restaurants — and sometimes across items on one menu — they don’t.

Source: Gordon Haskett Research Advisors analysis, reported by Restaurant Dive / Business Insider.

Why you can’t see it: the economics of not knowing

The reason this stays hidden isn’t laziness. It’s structural — and economists described the structure decades before delivery apps existed.

In 1970, George Akerlof published “The Market for ‘Lemons,’” the paper that won him a Nobel Prize and founded the economics of asymmetric information. His setup: a used-car seller knows whether a car is a lemon; the buyer can’t tell a good one from a bad one before paying. When one side knows something the other can’t observe, Akerlof showed, the market price stops carrying reliable information — and good outcomes get driven out by bad. A delivery menu is a small, everyday version of the same asymmetry. The restaurant knows its in-store price. You, looking at the app, do not. You can’t price what you can’t see.

George Stigler had already named the symptom in 1961, in the paper that launched search theory. Why does the same item sell at different prices at the same time? Because comparing prices costs effort, and people stop searching before they’ve checked everywhere. His line is the whole problem in a single sentence:

“Price dispersion is a manifestation — and, indeed, it is the measure — of ignorance in the market.”

George Stigler, “The Economics of Information,” Journal of Political Economy (1961)

Read that against a delivery order. The dispersion between the app price and the in-store price isn’t a sign your dish is better or different. It’s a measure of how hard it would be for you to check — and on a delivery app, checking is nearly impossible. You’d have to call the restaurant, or stand in it, to learn the real number. So you don’t. The markup survives precisely because the asymmetry is so cheap for the restaurant to maintain and so expensive for you to pierce.

Sources: Akerlof, “The Market for ‘Lemons,’” Quarterly Journal of Economics (1970); Stigler, “The Economics of Information,” Journal of Political Economy (1961).

Chick-fil-A got sued for exactly this

If you want proof that the invisibility is the offense — not the markup itself — look at what happened to Chick-fil-A. In 2023 the chain settled a $4.4 million class action in Georgia federal court. The accusation: it advertised free or cheap delivery — flat fees of $2.99 to $3.99 — while quietly raising its delivery menu prices 20 to 30% above in-store. An independent analysis pegged Chick-fil-A’s average delivery markup at 29.8%, against a fast-food average of 15.3%.

Marking up delivery prices wasn’t itself the problem; nearly every restaurant does it. As the coverage put it, where Chick-fil-A “went wrong” was “failing to disclose their prices while advertising their flat fee.” The remedy the settlement imposed is the tell: Chick-fil-A agreed to add a notice to its app and site stating that “product prices may be higher for delivery orders.” The fix for the harm was disclosure — closing the information gap. That’s Akerlof’s lesson, enforced by a court: when you can’t repair the asymmetry any other way, you make the hidden number visible.

The catch for your split: A disclosure that prices “may be higher” still doesn’t tell you how much higher, or which items, or how your dish compares to your friend’s. The asymmetry shrinks; it doesn’t close. You’re still dividing numbers you can’t fully audit.

Source: Food On Demand, “Chick-fil-A Settles $4.4M Lawsuit as Customers Demand Delivery Price Transparency” (2023).

What invisible, uneven markup does to a split

Now put the pieces together at the table. Two friends share one delivery order and split it the careful way — by item, so each pays for what they got. It looks airtight. Watch what the hidden markup does to it.

One delivery order, split by item
Sam — high-demand chain combo (app price)$18.00
↳ invisible markup (~29%)~$4.00
Riley — casual-dining side + drink (app price)$9.00
↳ invisible markup (~3%)~$0.26

Split by item, Sam pays $18 and Riley pays $9. Perfectly fair, right? Each paid for their own order. But Sam’s $18 carries roughly $4 of channel markup and Riley’s $9 carries about a quarter. Order the same food directly from the restaurants and Sam pays ~$14, Riley ~$8.74 — and the gap between them shrinks from $9.00 to $5.26. The delivery channel didn’t tax them equally. It marked up Sam’s choice at roughly ten times the rate it hit Riley’s, and neither of them could see it happen.

The split everyone trusts:
Sam $18.00  |  Riley $9.00  →  “we each paid for what we ordered”

What the channel actually charged for the same food:
Sam ~$4.00 markup  |  Riley ~$0.26 markup
Same receipt, wildly different hidden surcharge — invisible to both.

Now run the cruder version. If those two had just split the $27 order evenly — $13.50 each — Riley would have eaten half of Sam’s $4 markup on top of overpaying for the order itself. An even split doesn’t just ignore who ordered more. It socializes a markup that was never evenly earned.

Itemized splitting helps — but it can’t make it fair

Here’s the uncomfortable turn. Splitting by item is the right move — it beats an even split every time. But it can’t fully fix this, and it’s worth being honest about why.

When you split by item, each person pays their own dish’s app price — markup and all. That correctly stops one person from subsidizing another’s bigger or pricier order. What it can’t do is tell you that the markup rate differs from dish to dish and restaurant to restaurant. Sam still overpays by $4 relative to ordering direct; Riley by a quarter. “You pay for what you ordered” turns out to mean “you pay for a markup you couldn’t see, at a rate you didn’t choose.” Fair between the two friends, maybe. Fair compared to the food’s real price, no.

The only thing that would make it truly fair is the one thing the app withholds: the in-store price. Without it, every split — even and itemized alike — divides distorted numbers. The best you can do is keep the distortion from spreading to people who didn’t order it, and go in knowing the receipt isn’t telling you the whole truth.

Even splitSocializes one person’s markup across everyone — the least fair
Itemized splitKeeps each person’s markup on their own dish — fairer, still inflated
Order directRemoves the markup entirely — the only real fix

How to protect the split

You can’t see the markup, but you can keep it from quietly unbalancing who pays what. Four habits, in order of how much they help.

1

Order direct when the option exists

Many restaurants run their own pickup or delivery ordering at in-store prices. Skipping the third-party app skips the markup entirely — the single biggest lever. Same food, real prices, fair split.

2

Keep the group on one restaurant

When everyone orders from the same listing, the markup is at least uniform across the order. Mixing restaurants (or virtual brands) mixes markup rates, which is when an even split goes silently lopsided.

3

Spot-check one price before you commit

Pull up the restaurant’s own website for a single item you’re about to order. If the app price is wildly higher, you’ve just measured your ignorance — and you can decide whether the convenience is worth it.

4

Always split by item, never evenly

Even splitting takes the person with the most marked-up order and spreads their premium across everyone. Itemized splitting keeps each person’s hidden markup on their own plate. It’s not perfect, but it’s the fairest option the receipt allows.

The honest bottom line: No splitting method can show you a price the app refuses to show. What it can do is make sure the markup you can’t see lands on the person who chose it — not on the friend who ordered a side salad.

Where splitty fits

splitty doesn’t claim to undo any of this — it can’t see the in-store price any more than you can. What it does is read the printed delivery receipt and split it line by line, assigning each item to the people who actually shared it. The markup baked into a dish rides along with that dish, to the person who ordered it, instead of being smeared evenly across the table.

That’s the honest version of fair here: not “splitty reveals the real price” — it doesn’t — but “the friend who ordered the lightly marked-up salad doesn’t subsidize the friend who ordered the heavily marked-up combo.” Tax and tip distribute proportionally to each person’s share, so the inflated base at least gets divided in the right proportions. For the rest — the fees on top of the food — our delivery fees guide breaks down who pays what.

FAQ

Delivery markup and splitting questions

01 Do delivery apps really charge more than the restaurant?
Usually, yes — but not through a separate fee. Restaurants raise the menu prices on the app itself to offset the 15–30% commission the platform charges. Gordon Haskett found fast-food delivery prices average more than 15% higher than in-store, before delivery and service fees are added on top.
02 Why is the markup different at different restaurants?
Because platforms let restaurants choose from a range of commission rates, and restaurants have different margins and pricing power. A high-demand chain like Chick-fil-A marks up around 30%; a casual-dining chain might add just 3%. There’s no single ‘delivery tax’ — the rate is set restaurant by restaurant, and the app never shows it.
03 Is it fair to split a delivery order evenly?
It’s the least fair option. An even split spreads one person’s hidden markup across everyone, so whoever ordered the lightly marked-up item subsidizes whoever ordered the heavily marked-up one. Splitting by item keeps each person’s markup on their own dish — still inflated, but fair between the people at the table.
04 Can splitty show me the real in-store menu price?
No — and no splitting tool can. The app only shows the marked-up delivery price, so that’s all splitty can read. What splitty does is divide that printed receipt by item, so the invisible markup lands on whoever ordered the dish instead of being split evenly across the group.

Related guides

Delivery Fees Explained
Where every delivery, service, and small-order fee actually goes
DoorDash Fees in 2026
Why partitioned pricing makes a $16 dish check out at $30
Ghost Kitchen Confusion
When five ‘restaurants’ are one kitchen charging five prices

More in Practical Guides

All Guides

Sources

  1. 1 The Market for "Lemons": Quality Uncertainty and the Market Mechanism George A. Akerlof, The Quarterly Journal of Economics, 1970
  2. 2 The Economics of Information George J. Stigler, Journal of Political Economy, 1961
  3. 3 Food Ordering and Delivery: How Platforms and Restaurants Should Split the Pie Jaelynn Oh, Chloe Kim Glaeser, Xuanming Su, Management Science, 2023
  4. 4 Major QSRs hiking delivery menu prices by over 15% Gordon Haskett Research Advisors, Gordon Haskett Research Advisors, 2020
  5. 5 Chick-fil-A Settles $4.4M Lawsuit as Customers Demand Delivery Price Transparency Food On Demand, Food On Demand, 2023
  6. 6 Uber Eats raises marketplace fees Restaurant Dive, Restaurant Dive, 2026

Order what you want.

Three free bills, then $9.99 a year, or $24.99 once.