The $34 in charges nobody agreed to

A $62.40 group DoorDash order. Four friends. Everyone picked their meals, confirmed their items, closed the app. Then the receipt lands: $96.76. Delivery fee. Service fee. Express delivery somebody clicked by accident. Tax. Tip. $34.36 in charges that nobody discussed, nobody approved, and now nobody knows how to split. (The order is illustrative; the full receipt is below.)

This is the defining friction of modern group dining. Not the food. The fees. And unlike a restaurant check where one total gets divided, delivery orders stack several separate charges on top of each person’s food — charges that benefit the group unevenly, differ by platform, and vanish into a single payment on one person’s credit card.

2 in 3U.S. consumers who have used a food-ordering app (Purdue, Sept. 2024)
19%Average delivery-app menu premium over in-store prices, 25 brands (March 2023)
67%DoorDash and Caviar’s share of observed U.S. meal delivery sales (March 2024)

Sources: Purdue University Center for Food Demand Analysis and Sustainability, Consumer Food Insights (September 2024); Gordon Haskett Research Advisors, “Third-Party Delivery Menu Pricing Premiums: Spring 2023 Edition,” via Food On Demand; Bloomberg Second Measure, “Which company is winning the restaurant food delivery war?” (March 2024 data)

Why delivery bills are harder to split than restaurant checks

A restaurant bill has line items, tax, and tip. Three categories. A delivery order has food items with a menu markup you can’t see, then a delivery fee, a service fee, and depending on the order a small order fee, a regulatory response fee, an Express delivery fee, tax, and tip. Far more categories — and the person paying at checkout sees them scroll past in a blur of fine print.

The markup starts on the restaurant side. DoorDash’s merchant pricing page lists delivery commissions of 15%, 25%, or 30% of each order, depending on the plan a restaurant picks. Restaurants also price differently on the apps: Gordon Haskett Research Advisors found that 23 of the 25 restaurant brands it tracked charged more on delivery apps, with an average premium of 19% over in-store prices in March 2023. So before a single fee appears on your receipt, the food itself often costs more than it would in the restaurant.

Here is how that stacks up on an illustrative four-person order (New York City tax rate, DoorDash’s 15% service fee; the dollar amounts of the delivery and Express fees are examples).

4 meals (in-store price)$52.00
Menu markup (~20%)+$10.40
Delivery fee+$3.99
Service fee (15%)+$9.36
Express delivery fee+$2.99
Subtotal before tax and tip$78.74
Tax (8.875% of food)+$5.54
Tip (20% of food)+$12.48
Total charged$96.76

Before tax and tip, the four meals cost $52.00 at the counter and $78.74 delivered: 51% more once the markup and fees are stacked. That figure is our example’s own arithmetic. For a deep dive into every line item, see our complete delivery fee breakdown.

Sources: DoorDash, “Commission Rates & Pricing” (merchant plans); Gordon Haskett Research Advisors, Spring 2023 menu pricing premiums, via Food On Demand; DoorDash Help Center, “What fees do I pay?” (fee types; 15% service fee in markets including New York City). Receipt amounts are illustrative.

The psychology of invisible charges

Delivery apps sit on top of a specific psychological blind spot. Qingguo Ma’s team at Zhejiang University revisited the classic “pain of paying” in a 2024 PsyCh Journal study of mobile payment. Pain of paying explained the higher purchase intention with mobile payment only when prices were high. At low prices, the team proposed a second mechanism, a “pleasure of payment”, and tracked it through where shoppers’ eyes went: toward products’ benefits. The study tested product purchases, not delivery fees, but it shows that paying by phone changes more than the sting of paying.

“

We propose a new concept of “pleasure of payment” that is defined as an implicit and consumption-related hedonic response.

Ma, Tan, He, Cheng & Wang, PsyCh Journal (2024)

A 2025 study in Behavioral Sciences by Faraz and Anjum, a survey of 1,162 respondents, named a related effect “spendception”: the reduced resistance to spending when paying digitally. In their survey data, Spendception went with higher purchase behavior, and the authors write that digital payment systems “made buying feel less noticeable.”

Neither study looked at delivery fees; read them as background on how app payment feels. Richard Thaler’s 1999 review of mental accounting describes how people group spending into categories like housing and food, and sometimes hold each category to an implicit or explicit budget. A delivery fee has no category of its own. Our read: it is easy to leave out when each person pictures what their own dinner costs. When four friends add meals that come to $62.40, nobody is budgeting for the $34.36 that lands on top.

Sources: Ma et al., “Why does mobile payment promote purchases?” PsyCh Journal (2024); Faraz & Anjum, “Spendception,” Behavioral Sciences (2025); Thaler, “Mental Accounting Matters,” Journal of Behavioral Decision Making (1999)

The collective action problem of group delivery

Mancur Olson’s 1965 The Logic of Collective Action showed why shared costs are hard to collect in large groups: the free-rider problem. Many people who gain from a shared effort get the benefit even if they bear none of the cost. A group delivery order is a small version of that. In a two-person order, both people see the delivery fee. In a six-person order, it is easy to treat the fee as someone else’s problem.

Uri Gneezy, Ernan Haruvy, and Hadas Yafe measured a related cost in their 2004 restaurant field experiment. When diners knew the bill would be split evenly, they spent about 36% more per person than when each paid for their own meal. A group delivery order that the group plans to split evenly carries the same incentive: each extra item costs the person who adds it only a fraction of its price.

The key insight

Fee diffusion

Our own name for a pattern, with no study behind it: when one card pays and the split happens later, nobody at checkout sees their own share of the fees, so nobody owns them. The organizer is left holding the gap.

Then there is the person who fronts the order. They pay the full fee stack at checkout, then either absorb the difference or face the awkward task of itemizing every charge. And when something goes wrong — missing items, wrong orders — splitting refunds and credits adds another layer of complexity. Underneath it all is fairness. Xia, Monroe, and Cox’s 2004 review in the Journal of Marketing pulls together the research on how buyers judge whether a price is fair, and an organizer quietly covering everyone’s fees is paying a price nobody agreed to.

Sources: Olson, The Logic of Collective Action (1965), as summarized by Econlib; Gneezy, Haruvy & Yafe, “The Inefficiency of Splitting the Bill,” The Economic Journal (2004), mean meal cost 50.9 vs. 37.3 shekels; Xia, Monroe & Cox, “The Price Is Unfair! A Conceptual Framework of Price Fairness Perceptions,” Journal of Marketing (2004)

Platform-by-platform: what actually works

Every major delivery app handles group orders and split payments differently. DoorDash and Uber Eats can split a group order’s payment, but only if the organizer picks that option when setting up the order. Otherwise one card pays. And with ghost kitchens operating under multiple brand names from a single location, even figuring out what restaurant made your food can be confusing. Here is what each platform offers — and where it falls short.

DoorDash67% of observed U.S. sales (March 2024)

Group Orders: Yes. Each person adds their own items. DoorDash’s business FAQ says the admin can set a per-person budget limit.

Split Payment: “Everyone pays separately” option available. DoorDash’s business Group Orders FAQ says the delivery fee is then split evenly, the service fee in proportion to each order, and the Dasher tip is applied per participant. DoorDash documents the option on its Business support pages, so it may not appear on every account.

The catch: The creator has to choose it when starting the cart; to change it, they delete the cart and start over. Alcohol can’t go on a split-billing order. If the creator pays for everyone, the fees and tip fall on them.

Full DoorDash splitting guide | Group order walkthrough | DoorDash receipt guide

Uber Eats23% of observed U.S. sales (March 2024)

Group Orders: Yes. The host can set an optional deadline for everyone to pick their items and an optional spending limit.

Split Payment: “Guests pay for themselves” mode — participants pay their items, taxes, and fees through their own Uber Eats accounts. Uber says a split order can be split a maximum of 18 ways.

The catch: Everyone needs an Uber Eats account with payment info saved.

Full Uber Eats splitting guide | Uber Eats receipt guide

Grubhub8% of observed U.S. sales (March 2024)

Group Orders: Yes. On Grubhub for Business group orders, the host can set a budget for each guest. The host controls checkout.

Split Payment: Limited, and it depends on the account. Grubhub’s business guide says accounts with a “Split with Coworkers” feature can split the bill at checkout. Its Grubhub for Work migration FAQ says the Open Group Order owner pays for the cart and describes split payment as in development.

The catch: Grubhub’s business guide says the host is responsible for checkout, covering costs and fees on the entire order. Unless the account has Split with Coworkers, the host pays and collects afterward.

Grubhub receipt guide

The universal problem: On a regular order, or a group order where one person pays for everyone, that person absorbs delivery fees, service fees, and tip — then hopes the group Venmos them back. For the psychology of why those IOUs often go unpaid, see the Venmo Later problem.

Sources: DoorDash Business Support, “Group Orders FAQ”; Uber Help, “How to Place a Group Order”; Grubhub for Business, “Group Orders”; Grubhub, “FAQ - Grubhub for Work Migration”; Bloomberg Second Measure (March 2024 sales shares). Checked October 2026.

Built-in splitting vs. after-the-fact splitting

There are two approaches to splitting a delivery order. Each has clear tradeoffs.

Before ordering

App’s built-in group order

Everyone adds items to the same cart through a shared link. Payment is handled at checkout.

Everyone sees what they’re ordering
No need to track who ordered what
When one person pays for everyone, they front the fees
Everyone needs an account (Uber Eats)
DoorDash’s split divides the delivery fee by head
After ordering

Scan the receipt, split everything

One person orders and pays. Then the receipt gets split: each item to the people who ordered it, and the extra charges divided by order size or evenly, as the group agrees.

Works with any platform’s itemized receipt
Charges can follow each person’s order size
Only the person who orders needs a delivery account
One person fronts the cost temporarily

The after-the-fact approach puts the allocation rule in the group’s hands. When you order $22.40 worth of food and your friend orders $8, a proportional split gives you the larger share of the charges that scale with the order. The built-in options get partway there: DoorDash’s split billing divides the service fee by order size and the delivery fee evenly, and Uber Eats’ “Guests pay for themselves” has each person pay their own taxes and fees. Both have to be picked when the order is set up. Of the most common splitting methods, equal division and “just Venmo me” ignore order size, and manual math works but is slow on a receipt with this many lines. To do the fee math right now without any app, the free delivery fee split calculator takes each person’s order plus the fees and shows the proportional and even splits side by side.

Common delivery splitting scenarios

Different group contexts create different splitting challenges. Here’s how to handle each one.

Roommates

The weekly group order

Roommates ordering together regularly. Take turns being the organizer to distribute the fee burden. Or use a receipt scanner to split proportionally each time. In the example below, the roommate who orders least pays about $217 more a year on even splits of a weekly order than on proportional ones.

Office

The team lunch order

One person places the order for the group. When the company pays, expense the full amount. When coworkers split, screenshot the receipt and divide by what each person ordered. See our office lunch guide for navigating the politics.

Friends

The hangout delivery

Movie night. Game day. Super Bowl party. Someone pulls up DoorDash. The organizer shouldn’t absorb every fee and the tip because they happened to have the app open. Split the receipt after — every fee explained.

Mixed orders

Different apps, one hangout

When the group can’t agree on one restaurant and orders from multiple apps, each receipt needs separate splitting. This is where food hall splitting strategies apply — multiple vendors, one group, parallel receipts.

The delivery tipping problem

Tipping on delivery orders creates a secondary splitting debate. Purdue University’s September 2024 Consumer Food Insights survey of 1,200 U.S. consumers found that, on average, people say they tip between 10% and 19% on a food delivery order. But 15% say they tip less than 10%, and 14% say they don’t tip at all. Habits that far apart collide on a group order, where one person picks the tip for everyone. Split evenly, everyone pays a share of a tip only one person chose.

10-19%Average delivery tip consumers report
15%Say they tip less than 10% on delivery
14%Say they don’t tip on delivery at all

The fair approach: include the tip in the total and split proportionally. In our example, Alex ordered $22.40 of the $62.40 food total, so Alex covers 35.9% of the $12.48 tip, or $4.48. For platform-specific tipping norms, see our Uber Eats tipping guide and complete tipping guide for 2026.

Source: Purdue University Center for Food Demand Analysis and Sustainability, Consumer Food Insights (September 2024), as reported by Purdue Agriculture News (October 2024)

The math of proportional fee splitting

Equal fee splitting is fast. Proportional splitting tracks what each person ordered. Here’s why the difference matters.

Take the illustrative order from the receipt above. Alex gets a $22.40 entree. Jordan gets an $8 side. Sam and Pat each get $16 meals. Total food: $62.40. Total fees, tax, and tip: $34.36. If they split those charges equally, each pays $8.59. But Alex ordered 35.9% of the food and should pay 35.9% of the charges ($12.33). Jordan ordered 12.8% and should pay $4.41.

Equal split: Jordan pays $8.59 in charges on an $8 meal (107% of the meal)
Proportional split: Jordan pays $4.41 in charges on an $8 meal (55% of the meal)
Difference: Jordan pays $4.18 more with equal splitting

That $4.18 per order adds up. If the same order repeats weekly for a year, the person who orders least pays about $217 more with equal splitting than with proportional splitting. Asked in advance, most diners in Gneezy’s 2004 study preferred to avoid the even split: 19 of the 24 diners asked (80%) said, before ordering, that they would rather pay individually than split evenly.

The delivery fee exception: Some groups argue delivery fees should be split equally regardless of order size, since everyone benefits equally from delivery. That’s a valid position. The key is making the choice explicit rather than defaulting to whatever the app charges the organizer. See our guide to priority fee splitting for more on this debate.

How the research shaped splitty

Here is how we connect the science of fair splitting to splitty’s design for delivery orders. The arrows show our design reasoning; the studies did not test these features.

Mobile payment changes how spending feels (Ma et al., 2024)→Read the line items printed on the receipt, so the charges are in front of everyone
When one person pays for the group, they absorb every fee→Tax and tip divided by what each person ordered
People sort spending into mental accounts (Thaler, 1999)→Items start split among everyone; tap to remove a person from what they didn’t share
Large groups invite free-riding (Olson, 1965)→A payment request for each person’s share, through the payment app they prefer
Buyers judge whether a price is fair (Xia et al., 2004)→Fair, not equal: each item goes to the people who shared it