Restaurant Profit Margins: Why Every Restaurant Is Now Two Businesses

Last updated August 31, 2026.

The modern restaurant is no longer one operating model. It is two businesses sharing one kitchen.

One business sells hospitality: a seat, a server, an atmosphere, and a meal experienced in the dining room. The other sells fulfillment: accurate food, packed well, ready on time, and handed to a customer or driver with as little friction as possible.

That split matters because the two businesses create revenue in very different ways. They also put different pressure on labor, menus, kitchen capacity, technology, and restaurant profit margins.

In “The restaurant business is changing beyond recognition”, published July 30, 2026, The Economist examines how technology, inflation, delivery, automation, and changing customer habits are reshaping restaurants. The clearest lesson for independent operators is practical: dine-in and off-premise demand can no longer be managed as if they were the same business.

What 157,240 Restaurant Orders Reveal

We analyzed one year of anonymized transaction data from the ten highest-volume restaurants on Avocado POS, ranked by paid net sales. The period runs from September 1, 2025 through August 31, 2026.

Across those restaurants, the data included:

• 157,240 paid orders

• $3.88 million in net sales

• $24.68 overall average order value

The channel split is where the two-business model becomes visible.

• Dine-in: 17.7% of orders but 41.3% of net sales, with a $57.54 average order value.

• Pickup, takeout, and delivery: 82.3% of orders and 58.7% of net sales, with a combined average order value of about $17.59.

In other words, the dining room produced fewer tickets but much larger checks. Off-premise fulfillment produced most of the order volume, but each order was worth much less on average.

The mix also varied dramatically by restaurant. The median restaurant recorded 85.0% of orders as pickup, takeout, or delivery, while individual restaurants ranged from just 0.6% to 100%. There is no single “normal” restaurant model inside this top-ten group. Some operators are primarily hospitality businesses. Others are high-throughput food fulfillment businesses. Many are both at once.

Business One: The Dining-Room Hospitality Engine

A dine-in order is not simply a takeout order with a table attached. It has its own economics.

The $57.54 dine-in average order value in this dataset was more than three times the $17.55 pickup and takeout average. That larger check can absorb front-of-house labor and occupancy costs, but only when the restaurant protects the experience that earns it.

The hospitality business depends on:

• Table turns without making guests feel rushed

• Servers who can guide choices and increase attachment rates

• Accurate coursing and dependable ticket times

• A dining room that gives customers a reason to stay

• Add-ons such as drinks, appetizers, and desserts

For this side of the restaurant, labor is part of the product. Cutting service indiscriminately may lower payroll for a week while weakening the experience that supports a higher check.

Business Two: The Off-Premise Fulfillment Engine

Pickup, takeout, and delivery reward a different system. The customer is buying certainty: the correct order, at the promised time, in packaging that travels well.

Because off-premise tickets were smaller in this dataset, the operating model has to favor speed and repeatability. Every unnecessary touch matters more. A confused pickup shelf, an unavailable modifier, a missed sauce, or five minutes of avoidable kitchen delay can erase the convenience the customer chose.

The fulfillment business depends on:

• A menu designed for travel and fast production

• Simple modifiers and fewer error-prone choices

• Clear prep, packing, and handoff stations

• Real-time item availability

• Reliable pickup and delivery quotes

• Direct customer relationships that support repeat orders

This is why a single menu does not always deserve a single operating plan. The best dine-in item may be a poor delivery item. A meal that sells atmosphere in the dining room may arrive tired after 25 minutes in a container.

For a deeper look at channel-specific item economics, see our guide to restaurant menu engineering.

Why Treating Both Businesses the Same Shrinks Restaurant Profit Margins

1. One Menu Can Hide Two Contribution Margins

Food cost is only the start. Each channel adds its own labor, packaging, payment, and service costs. An item with a healthy dine-in margin can become weak off-premise after packaging and fulfillment are included.

Track contribution margin by channel and by item. If a dish travels poorly, requires expensive packaging, or slows the line, redesign it, reprice it, or remove it from the off-premise menu.

2. One Kitchen Can Have Two Rushes at the Same Time

A full dining room and a burst of online orders compete for the same cooks, equipment, and expo capacity. Revenue can rise while guest satisfaction and restaurant profit margins fall.

Set realistic channel capacity. Adjust quoted pickup and delivery times when the kitchen is saturated. If needed, throttle off-premise demand during the most profitable dine-in window rather than accepting every ticket and disappointing everyone.

3. One Labor Plan Can Misread the Work

Dining-room volume needs hosts, servers, bussers, and hospitality. Off-premise volume needs production, packing, labeling, and handoff discipline. Scheduling only from total sales can hide where the work actually occurs.

Build labor plans from expected order mix, not revenue alone. A shift with 100 small pickup orders may require different staffing than a shift with 30 high-value tables, even when total sales are similar.

4. One Customer Record Can Become Somebody Else’s Asset

Third-party marketplaces can generate demand, but they often stand between a restaurant and its customers. Direct ordering gives the operator a better chance to build loyalty, understand repeat behavior, and market responsibly to customers who have opted in.

Our comparison of third-party delivery and restaurant-controlled delivery explains the broader tradeoffs.

Delivery Should Add Convenience, Not Take a Percentage of the Menu

Delivery is a small but instructive part of the Avocado top-ten data. Five of the ten restaurants recorded delivery orders during the period. Across 244 delivery orders, the average order value was $40.36 and the average delivery fee was $6.65.

That sample is too small to generalize to the entire restaurant industry, but it highlights an important design choice: the delivery cost should be visible and separate from the restaurant’s menu revenue.

With Avocado’s delivery feature, restaurants keep 100% of their menu price. The customer pays the quoted delivery fee. Avocado does not take a delivery marketplace commission from the menu price. Standard payment processing, taxes, refunds, and other applicable charges still apply.

That structure makes the economics easier to understand. The menu can be priced around food, labor, and restaurant margin. The delivery fee pays for delivery. Operators can offer convenience without quietly surrendering a percentage of every item sold.

A Weekly Scorecard for the Two-Business Restaurant

You do not need two buildings or two brands. You need two clear views of performance.

Review these metrics by dine-in, pickup and takeout, and delivery every week:

• Order count and sales mix: Where is demand moving?

• Average order value: Which channels produce larger baskets?

• Contribution margin: What remains after food, channel labor, packaging, discounts, and channel-specific costs?

• Ticket time: Is one channel slowing another?

• Refund and error rate: Where are mistakes most expensive?

• Repeat purchase rate: Which channels create customers who return?

Then make one operational change at a time. Simplify three off-premise modifiers. Move the pickup shelf. Change delivery hours. Train servers on two profitable add-ons. Pause a low-margin item during the rush. Small channel-specific improvements compound.

The Restaurant Is Changing, but the Operator Can Still Own the Model

The Economist is right that the restaurant business is changing quickly. The useful response is not to chase every technology or abandon hospitality. It is to recognize that hospitality and fulfillment now coexist, then design each one deliberately.

The top Avocado restaurants in this analysis show the split clearly: dine-in generated 41.3% of net sales from only 17.7% of orders, while pickup, takeout, and delivery generated most of the volume. Those channels share a kitchen, but they do not share the same economics.

Restaurants that measure the two businesses separately can protect restaurant profit margins without forcing every customer into the same experience. Make the dining room worth visiting. Make off-premise ordering worth repeating. Keep control of the menu, the customer relationship, and the economics of every order.

Methodology: Avocado analyzed anonymized paid-order records from the ten restaurants with the highest paid net sales on Avocado POS between September 1, 2025 and August 31, 2026. Net sales equal order subtotal less refunded subtotal. “Off-premise” in this article means orders marked for pickup, takeout, or delivery in fulfillment records; it does not prove where the meal was ultimately consumed. Restaurant identities were excluded. The results describe this cohort and should not be treated as a representative sample of all U.S. restaurants.