Restaurant industry statistics 2026: 30+ key numbers, sourced

Short answer: the restaurant numbers most often repeated are wrong, and the real ones are more useful. Restaurants do not fail at 90% in year one (real first-year closure measurements run from under 1% to 17% depending on study and year); the industry nets 3–5% on average with bars at 10–15% and cafés nearer 2.5%; prime cost belongs under 65% of sales; delivery marketplaces take a published 15–30% that often lands at 30–40% all-in; and internal theft quietly consumes about 4% of sales in unmanaged venues. This page collects the statistics we cite across every guide on this blog, each with its source and a link to the deep dive. Bookmark it, and steal it for your business plan.

Survival: the 90% myth versus the data

The honest reading: hospitality is genuinely risky, and roughly half of venues do not see year six, but the odds are nothing like the folklore, and the failures cluster around known causes: undercapitalisation, rent set above the 8% occupancy test, and owners flying without weekly numbers. The business plan guide shows how lenders actually read these risks.

Margins and cost structure

Delivery and channel economics

Channel mix is the least visible margin lever in the industry: identical venues with identical sales can net wildly different profits purely on where orders arrive. The full rate breakdown and a savings calculator live in the delivery commissions guide, and the strategy in delivery strategy.

Buying, franchising, wholesale

Loss, seasonality, and the quiet leaks

The New Zealand corner

Deep dives for each: hiring and wages, GST, opening costs, and licences.

How to use (and how not to misuse) these numbers

• Benchmarks are diagnostic ranges, not targets: a café at 4% net is beating its format; a bar at 4% is leaving half its potential on the table.
• Ranges reflect real variance between sources and venues; if a citation needs one number, take the middle and name the range.
• Time-stamp what you borrow: commissions and wage rates in particular move yearly, and this page states its year.
• Every section links its working; cite the deep dive when the claim carries weight.

Frequently asked questions

What percentage of restaurants fail in the first year?
Far fewer than folklore says. The famous “90% fail in year one” line traces back to an unsourced TV commercial and has been debunked repeatedly: a UC Berkeley study measured about 17% first-year closures, the National Restaurant Association’s long-run average failure figure is around 30%, and Datassential measured US first-year closures at just 0.9% in 2025. Roughly half of restaurants are still trading after five years. Risky, yes; a coin-flip against you nine times over, no.

What is the average restaurant profit margin?
Industry-wide, roughly 3–5% net. By format: full-service restaurants 3–6%, quick service 6–10%, cafés around 2.5% on average with a wide spread, bars 10–15%, food trucks 6–9%. The spread inside each format is bigger than the spread between formats, which is why operators compare themselves to their own format’s band and then hunt the one cost line that explains their gap.

How much do delivery apps take from restaurants?
Published tiers run 15–30% of the order: DoorDash lists 15/25/30% plans (6% for pickup), and Uber Eats moved its entry tier from 15% to 20% in early 2026, with surcharges pushing effective rates on many orders to 30%. Once promotions, ad placements and assorted fees are counted, many venues measure a true 30–40% cost per marketplace order, which usually exceeds the entire net margin of the dish.

What should a restaurant’s prime cost be?
Under 65% of sales, with strong operators nearer 60%. Prime cost is food and beverage cost (typically 28–35% of sales) plus total labour (25–35%), and it is the single best weekly predictor of profitability because it captures the two costs management can actually move. Above 65% for more than a couple of weeks, the bottom line is almost certainly negative, whatever the till says.

How much revenue do restaurants lose to theft?
Industry loss-prevention estimates put internal theft around 4% of restaurant sales, up to about 7% in quick service, and attribute roughly 75% of inventory shrinkage to theft rather than genuine waste. Against a 3–9% net margin, that means an unmanaged venue can lose most of its profit line to shrinkage without a single dramatic incident, which is why counting stock and POS controls pay like a marketing campaign.

Where do these statistics come from?
Each figure on this page names its source in the table beside it: platform rate cards for commissions, MBIE for NZ wage rates, JLL’s 2026 fit-out guide for build costs, the UC Berkeley and Datassential studies for survival rates, and industry association estimates for margins and shrinkage. Ranges are quoted where sources vary. Statistics move; where a number is time-sensitive we date it, and the deep-dive guide linked beside each section holds the full working.