Insights

Restaurant Profit Margin Benchmarks (2026 Data Study)

The 2026 benchmark study: net margin, prime cost, food and labor cost, and revenue per seat by segment, plus how a 30% commission collapses a thin margin toward zero.

PA
Pankaj Avhad
Jul 11, 2026ยท13 min read
Share:
Insights
Conversion24%
Retention67%
AOV$42

TLDR

The average restaurant net profit margin in 2026 is 3% to 9%, built on a prime cost (food plus labor) of 55% to 65%, food cost of 28% to 35%, and labor of 25% to 35%. Quick-service runs highest at 6% to 12%, casual dining the thinnest at 3% to 7%. Because the margin is so thin, a 25% to 30% third-party commission on a $20 order collapses roughly $3.00 of take-home to about $0.50, which is why the channel mix decides profitability more than the menu does.

TLDR: The average restaurant net profit margin in 2026 is 3% to 9%, built on a prime cost of 55% to 65%, food cost of 28% to 35%, and labor of 25% to 35%. Quick-service runs highest, casual dining the thinnest. This is a reference study you can cite: net margin, prime cost, and revenue per seat by segment, in charts. Then the part most benchmark reports skip: because the margin is so thin, a 25% to 30% third-party commission on a $20 order collapses roughly $3.00 of take-home to about $0.50. The channel your orders flow through decides profitability more than the menu does.

The 2026 restaurant benchmark snapshot

3% to 9%
Net profit margin
What lands in the owner's pocket after everything
55% to 65%
Prime cost
Food plus labor combined, the number to watch
28% to 35%
Food cost
Cost of goods sold as a share of revenue
25% to 35%
Labor cost
Wages, taxes, and benefits as a share of revenue
65%Prime cost above 65% is the danger line. Most restaurants that cross it stop making money.

Sources: NRA 2026 State of the Industry, Restaurant365, Toast, Square.

The 2026 restaurant profit benchmark, in one snapshot

The average full-service restaurant keeps 3% to 9% of revenue as net profit after every cost is paid. That is the headline number, and it is thin enough that a single bad line item can wipe out the entire year's profit. To read the rest of this study, you need four numbers and one threshold:

  • Net profit margin: what the owner keeps after all costs. Industry range 3% to 9%.
  • Prime cost: food cost plus labor cost, combined. Target 55% to 65%.
  • Food cost: cost of goods sold as a share of revenue. Range 28% to 35%.
  • Labor cost: wages, payroll taxes, and benefits as a share of revenue. Range 25% to 35%.
  • The 65% line: prime cost above 65% of revenue is where profit usually turns negative, because occupancy, insurance, marketing, and other fixed costs still have to be paid out of what is left.

These ranges come from the NRA 2026 State of the Industry report, Restaurant365 operator benchmarks, and Toast and Square transaction data. They describe a healthy independent restaurant. The rest of this article breaks them down by segment, then shows what happens to every one of these numbers the moment an order routes through a marketplace app instead of a channel you own.

If you want the survival context behind these margins, our companion study on the real restaurant failure rate covers year-by-year odds and why 42% of operators entered 2026 unprofitable. This post is the money-map that sits underneath those odds.

How to read the numbers (definitions before benchmarks)

A benchmark is useless if you compare the wrong line. Three definitions keep operators honest.

Net profit margin is not gross margin. Gross margin (revenue minus food cost) can look healthy at 68% while the business loses money, because labor, rent, and everything else come out of that gross figure. When this study says "margin," it means net margin: the last line, the amount that actually belongs to the owner.

Prime cost is the number that predicts survival. Food and labor are the two largest and most controllable costs in any restaurant, so operators combine them into prime cost and watch it like a fuel gauge. At or below 60% you have room to absorb a slow week. At 65% you are running on fumes. Above 65% the math stops working, no matter how good the food is.

Revenue size changes everything. A restaurant doing $2 million absorbs its fixed costs across far more covers than one doing $500,000, so the larger operation posts a materially higher margin on the same menu. When you benchmark yourself, compare against restaurants in your revenue band and your segment, not against the industry average, which blends a bar's 12% with a cafe's 3%.

Net profit margin by segment

Net margin varies more by service model than by anything else. Counter-service concepts keep labor low and turn covers fast, so they sit at the top. Full-service casual dining carries the heaviest combined load of food quality, table service, and occupancy, so it sits at the bottom.

Net Profit Margin Range by Segment (2026)

Typical net margin after all costs. Ghost kitchen shown at its ~6% midpoint band.

Sources: NRA 2026 State of the Industry, Restaurant365 benchmarks, Toast, Square.

SegmentTypical net marginWhy it lands here
Quick-Service (QSR)6% to 12%Low labor, high volume, fast table turns
Fast Casual5% to 10%Premium ingredients offset by limited service
Fine Dining3% to 8%High check averages, but food and skilled labor cost the most
Ghost Kitchen~6%No front-of-house or dining room, but heavy delivery-channel exposure
Casual Dining3% to 7%Full service plus occupancy on every square foot

The ghost kitchen number carries an asterisk that matters for this whole study. A delivery-only kitchen strips out the dining room and most front-of-house labor, which should push margin up. It lands at only about 6% because nearly every order arrives through a third-party marketplace that takes a commission. The cost structure is lean, but the channel structure is expensive. Hold that thought, because it is the pivot of this entire report.

Prime cost breakdown by segment

Split prime cost into its two parts and the segment differences get concrete. Food cost is relatively stable across the industry at 28% to 35%. Labor is where the models diverge: a quick-service counter needs a fraction of the staff a fine-dining room does.

Prime Cost by Segment: Food + Labor

The 65% line is the viability threshold. Cross it and profit usually turns negative.

Food cost 28% to 35% and labor 25% to 35% are the industry ranges. Sources: NRA, Restaurant365, Toast.

SegmentFood costLabor costPrime cost
Ghost Kitchen33%20%53%
Fast Casual30%27%57%
Quick-Service30%28%58%
Casual Dining32%32%64%
Fine Dining34%33%67%

Notice fine dining crossing the 65% line. That is not a failing restaurant, it is the reality that white-tablecloth service and premium ingredients push both components high at once. Fine-dining operators earn their margin back through check average, wine and beverage sales at much higher food-cost efficiency, and disciplined labor scheduling. The lesson for every segment: if your prime cost sits near or above 65%, your beverage program, your menu engineering, and your labor scheduling have to be excellent, because you have no cushion left for waste.

For the tactical side of pulling these percentages down, our guide on how to increase restaurant sales covers menu engineering, upselling, and the retention plays that raise revenue without raising prime cost.

Revenue per seat economics

Prime cost tells you whether each dollar is efficient. Revenue per seat tells you whether you are generating enough dollars in the first place. It is a simple division: total sales divided by the number of seats.

A 40-seat casual restaurant doing $1.2 million a year generates $30,000 per seat annually, or about $82 per seat per day. Whether that is good depends entirely on your turns. If those 40 seats fill twice per dinner service at a $28 average check, you are leaving capacity on the table compared to a restaurant turning the same seats three or four times. Revenue per seat is the fastest way to compare a 30-seat cafe against a 120-seat family restaurant on equal footing, because it normalizes for size.

Pair the two metrics and you get a clean diagnostic:

  • High revenue per seat, prime cost under 60%: the profitable-operator signature. Protect it.
  • High revenue per seat, prime cost over 65%: you are busy but leaking. The problem is cost control, not demand.
  • Low revenue per seat, prime cost under 60%: disciplined but underfed. The problem is demand and channel reach, not the kitchen.

That third case is the most common trap for independents, and it is exactly where the channel mix decides the outcome. Which brings us to the number that quietly overrides all the benchmarks above.

The margin killer: commission drag

Here is the finding that should reshape how you read every benchmark in this study. A restaurant profit margin of 3% to 9% is not just thin, it is smaller than a single third-party commission. So when an order routes through a marketplace app that charges 25% to 30%, the fee is not a haircut on your profit. It is larger than your entire profit on that order.

Where a $20 Order Goes: Direct vs Marketplace

The same order nets $3.00 direct and about $0.50 after a 25% to 30% marketplace fee.

Illustrative $20 order with typical 30% food and 20% labor loads. Marketplace fee assumes a menu markup to offset commission. Sources: NRA, DirectOrders analysis.

Follow a $20 order through the costs. Food takes about $6.00. Labor takes about $4.00. Overhead (rent, utilities, insurance, packaging, processing) takes roughly $7.00. That leaves about $3.00 of take-home on a direct order, which is a healthy 15% contribution margin at the order level.

Now route the identical order through a marketplace at 25% to 30%. Even after raising the in-app menu price to offset part of the commission, the fee and passed costs pull take-home down to roughly $0.50. Same food, same kitchen, same customer. The channel took five-sixths of the profit.

This is why the ghost kitchen margin was stuck at 6% despite the leanest cost structure in the industry. It is not a kitchen problem, it is a channel problem. And it is the same trap independents fall into when marketplace apps drift from a discovery tool to the majority of their volume. We break down the true marketplace rates and the break-even math in the real math of restaurant commission, and we unpack the fees hidden inside "commission-free" pitches in the hidden cost of zero commission platforms.

What one point of commission is worth, compounded

A single order looks survivable at $0.50. The damage is in the volume. Take a restaurant doing 500 online orders a month at a $20 average order. That is $10,000 of online revenue, the same food cost, the same labor, either way. The only variable is the channel.

Cumulative Profit on the Same 500 Orders/Month

Direct ordering clears the $249 flat fee inside week one, then compounds all year.

Assumes 500 orders/month, $20 average order, $3.00 direct take-home vs $0.50 after marketplace fees. Illustrative.

At $3.00 take-home per direct order, that channel contributes about $1,500 a month. On a flat-fee direct platform like DirectOrders at $249 a month with zero commission, the platform cost is covered by the first few days of orders, and the restaurant keeps roughly $1,251 a month net. Route the same 500 orders through a marketplace at $0.50 take-home and the channel contributes about $250 a month.

ChannelTake-home per $20 order500 orders/monthOver 12 months
Direct (flat $249/mo, zero commission)$3.00~$1,251 net~$15,000
Marketplace app (25% to 30% fee)$0.50~$250~$3,000

That is a $12,000 annual swing on identical gross sales. Nothing about the food, the labor, or the demand changed. The restaurant simply kept its own margin instead of renting it out one order at a time. On the profit-margin math from the top of this study, $12,000 recovered is often the difference between a 3% year and a 6% year.

The repair path: flat fee plus direct restores the margin

The fix is not to abandon marketplaces overnight, they still deliver discovery. The fix is to change the mix so your highest-intent, repeat customers order through a channel you own, where the commission is zero and the fee is flat.

  • Zero commission, ever. No per-order fee means the take-home on that $20 order stays at $3.00 instead of $0.50. You keep 100% of every sale.
  • A flat, predictable cost. DirectOrders Pro is $249 a month and Pro plus Voice AI is $349 a month (500 voice minutes included). The cost does not grow with your success, which is the opposite of a commission.
  • The channels to actually capture direct orders. A branded website and online ordering, 15+ ordering channels, and AI voice ordering that answers the phone 24/7 so the calls that come during a dinner rush turn into orders instead of voicemails.
  • Cash flow that matches the margin. Same-day payouts and same-day dispatch through Uber Direct and DoorDash Drive, so a thin-margin business is not financing a week of float.
  • No lock-in. No annual contract, cancel anytime, POS sync with Square, Toast, and Clover, and you own your customer data. Live in 2 hours or we white-glove you free.

See the full breakdown on the pricing page. The point is not the feature list, it is the arithmetic: on a 3% to 9% margin, moving orders from a 25% to 30% channel to a 0% channel is the single highest-return operational change available to an independent restaurant.

The levers operators actually control

Benchmarks are a mirror, not a destiny. Every number in this study has a lever attached to it.

Metric2026 benchmarkThe lever
Food cost28% to 35%Menu engineering, portion control, supplier terms, waste tracking
Labor cost25% to 35%Scheduling to demand, cross-training, cutting no-show and phone chaos
Prime cost55% to 65%Watch it weekly, not monthly; keep it under 60%
Channel mixvariesShift repeat orders from 25% to 30% marketplaces to a 0% direct channel
Revenue per seatsegment-dependentTurn tables faster, raise check average, extend dayparts

Food and labor get the most attention because they are the biggest lines. But they are also the hardest to move a full point on without hurting the guest experience. The channel-mix lever is different: it recovers full points of margin without touching the food, the staff, or the price the loyal customer sees. That is why it belongs at the top of the 2026 priority list for any operator whose marketplace share has crept past a third of online volume.

Sources and methodology

This study synthesizes published 2026 restaurant financial benchmarks with an order-level commission model.

  • Margin, prime cost, food, and labor ranges: NRA 2026 State of the Industry report, Restaurant365 operator benchmarks, and aggregated Toast and Square transaction data. Segment ranges reflect typical independent operators, not chains, which post different structures at scale.
  • Segment prime cost figures: midpoints within the published industry ranges, adjusted for the labor profile of each service model (counter service versus full service versus delivery-only).
  • Commission drag model: an illustrative $20 order carrying a typical 30% food load and 20% labor load, with a 25% to 30% marketplace fee net of a partial menu markup. The cumulative-profit projection assumes 500 online orders per month at a $20 average order.
  • Pricing figures: DirectOrders published plans, Pro at $249 per month and Pro plus Voice AI at $349 per month, zero commission.

Benchmarks describe the middle of the distribution. Your restaurant is a data point, not the average, so use these ranges to locate yourself, then work the lever attached to your weakest line. For the survival odds behind the margins, read the restaurant failure rate study. For the revenue plays that widen the margin from the top, read how to increase restaurant sales.

Frequently Asked Questions

The average restaurant net profit margin in 2026 is 3% to 9%, depending on segment. Quick-service restaurants run highest at roughly 6% to 12%, fast casual at 5% to 10%, fine dining at 3% to 8%, ghost kitchens around 6%, and full-service casual dining the thinnest at 3% to 7%. Net margin is what remains after food, labor, occupancy, and all other costs, so a $500,000 restaurant at a 5% margin keeps about $25,000 for the year.

Related resources

Related Articles

17%visible
CITATION TRACKER
ChatGPT cited
Perplexity cited
Gemini miss
Copilot cited
Claude miss

83% of restaurants are invisible in ChatGPT (Local Falcon, 2026)

Insights

How to Track Your Restaurant's AI Search Visibility (2026): Tools, Manual Methods, and What to Measure

83% of restaurants are invisible in ChatGPT (Local Falcon, 2026). Here is the operator-grade playbook for measuring whether ChatGPT, Perplexity, AI Overviews, Claude, and Copilot actually cite your restaurant: 7 metrics, 13 query buckets, 12 tools, and a free GA4 setup that catches the 70% of AI traffic that hides as Direct.

PA

Pankaj Avhad

May 6, 2026ยท15 min read
COMMISSION-MATH.XLSXFounder ledger
1DoorDash Premier30%
2Uber Eats blended26%
3Grubhub all-in25%
4DirectOrders$249/mo
NET SAVINGS / YR$43K-$45K
26.4FTC MEDIAN
Insights

The Real Math of Restaurant Commission: A Founder's Open Spreadsheet (2026)

The DirectOrders founder's open spreadsheet on the real marketplace commission rates restaurants actually pay (FTC median 26.4%), the break-even formula for going direct, and a modeled 12-month switch on an $80K/month restaurant.

PA

Pankaj Avhad

May 6, 2026ยท22 min read
Open
1,000
Yr 1
830
Yr 5
520
Yr 10
350
Yr 15
200
90% myth = false
0% survive yr 1
Insights

Restaurant Failure Rate (2026): Real Data, Real Odds, and How to Win

The 90% restaurant failure myth is wrong. See the real survival rates by year, the profit margins that predict who lasts, and what the 20% who make it past 15 years do differently.

PA

Pankaj Avhad

Apr 14, 2026ยท18 min read

Topics:

restaurant profit margin benchmarks 2026restaurant profit margins by typerestaurant prime cost benchmarkaverage restaurant profit marginrestaurant revenue per seatrestaurant food cost percentagerestaurant labor cost percentagerestaurant financial benchmarksrestaurant commission dragrestaurant profitability

Ready to grow your direct orders?

See how DirectOrders can help your restaurant keep more revenue and own your customer relationships.