Masterestaurant Analysis of restaurant paid advertising 2026: the 70% that decides where every dollar lands

Finding 1 — Seventy percent want to order from you, not from the app: what that means for your budget

Seventy percent of diners prefer to order directly from the restaurant and only 46% prefer third-party apps, according to Lightspeed (Online Ordering Statistics, 2025), and that 24-point gap is the first media decision almost nobody makes. If your ad lands on a DoorDash or Uber Eats profile, you are buying traffic and handing it to a channel that charges between 15% and 30% commission per order (Rezku, 2026) and that, once surcharges are counted, reaches an effective 35%-45% of the ticket (CloudKitchens, 2026). Pay to bring the diner to YOUR domain, your phone, your ordering system. That same Lightspeed report records that 42% use the apps solely to reorder something they already know: that diner is already yours, and yet you are renting them back from a third party month after month. Change the click destination before you touch a single image in the ad.

Finding 2 — Where the ad sends people matters more than the creative

A US$40 order coming through an app leaves you, at the effective 35%-45% commission documented by CloudKitchens (2026), somewhere between US$22 and US$26 of real revenue, while the same order through your own channel leaves the full US$40 minus the payment gateway. If your cost per acquired order is US$6, on your own channel that is 15% of the ticket, and through the app it goes past 25% of what you actually collect. Same creative, same audience, same budget, and profitability splits in two because of a link decision that takes thirty seconds. Anyone who has never run that differential is not optimizing advertising: they are funding the app's growth. Impressions, reach, engagement and cost per click never touch the register, and they are precisely the four that decide whether the owner raises or lowers the budget each month. At Masterestaurant the frame is different and it is not up for negotiation: paid advertising is a variable acquisition cost competing with food cost, payroll and rent for the same contribution margin.

Finding 3 — The four metrics that govern media spend and never show up in the P&L

Diego F. Parra reduces it to a rule you can audit on one sheet: if the average ticket cannot absorb the cost of acquiring that diner plus the cost of serving them, the ad stopped being marketing and became a subsidy to the diner paid out of EBITDA. With a 32% food cost —the acceptable ceiling per dish, never the target— a US$25 ticket leaves US$17 of gross margin, and that is where acquisition cost gets decided, not in the CPC. Before you put another dollar into media, look at your reviews: 96% of consumers are willing to write one and 89% expect a reply to both positive and negative reviews, per BrightLocal's Local Consumer Review Survey 2025. That 89% is operational work, not media budget, and it acts on exactly the same diner your ad just delivered. Restroworks (Restaurant Social Media Statistics, 2025) further documents that 62% check the restaurant's page before deciding.

Finding 4 — The review, and the reply, are worth more than the paid push

Translated into operations: your ad pushes the diner toward a page and a set of reviews you have not curated, so you are paying to expose your weakest point. Half an hour a day of a manager answering reviews costs less than a 10% budget increase and moves conversion on the very same traffic. Eighty-one percent of consumers would join a loyalty program if offered one (Businessdasher, 2025) and 78% say they are more likely to visit a restaurant where they earn points, according to the National Restaurant Association's State of the Restaurant Industry 2025. A restaurant that advertises without capturing the diner's data pays for acquisition every single time, forever. Add that nearly 90% of consumers would use app-exclusive offers (National Restaurant Association 2025, via Lightspeed) and the arithmetic changes shape: the first paid order no longer has to repay the entire acquisition cost, because a second and a third visit arrive with no CPC attached.

Finding 5 — Loyalty and first-party data: why the second visit should not cost you another click

Media spend stops being a tap you open and close each month and becomes the front door to a base you control. A discount applied at the wrong hour destroys margin; applied to the valley it builds margin. Restroworks (Restaurant Coupon Statistics), citing PepsiCo Partners 2025, records that for 62% of consumers time-based offers raise the likelihood of a visit, and that 40% attend happy hour weekly. There sits the difference between running a 20% discount on a Friday at 8:00 p.m., when your dining room is already full and you just gave away a fifth of the ticket to someone who was coming anyway, and running that same 20% for a Tuesday between 3:00 and 6:00 p.m., when payroll is already paid and every extra cover contributes almost clean margin. Savings.com 2025, in the same Restroworks compilation, adds that 82% say coupons and discounts help them against high prices: the instrument works, the daypart decides whether it costs you or pays you.

Finding 6 — SMS, gift cards and BOGO: the channels your media spend should be feeding

Eighty-four percent of consumers opted in to receive texts from at least one business in 2025, according to Sakari, and that permission is the real asset your paid advertising ought to be buying. Compare the orders of magnitude: an ad needs budget every time it wants to appear, while an SMS to your own base costs cents per send. Along the same line, Capital One Shopping (Gift Card Statistics, 2026) reports that 52% of consumers buy restaurant gift cards and that 61% spend beyond the card's value, with an average overspend of US$31.75 —money collected up front that funds the media itself—. Ninety-three percent have used a BOGO offer at least once (Capital One Shopping 2025, via Restroworks), so the mechanic needs no invention: it needs a destination of your own where the diner leaves a phone number. Suppose tomorrow you lose 50% of your media budget.

Finding 7 — What I would do if tomorrow your ad budget were cut in half

The restaurant that only bought clicks into an app loses half its incremental orders within fourteen days, with nothing to cushion the fall. The one that had been capturing data keeps its SMS base —recall the 84% opt-in measured by Sakari (2025)—, its points program with that 78% propensity reported by the National Restaurant Association (2025) and its answered reviews against an 89% that expects them (BrightLocal, 2025); it loses new reach, not the business. That is the stress test separating a healthy ad account from a dependency. Open your platform today, filter the campaigns whose destination is a third-party profile and redirect that budget to your own ordering channel: the result gets measured in next month's contribution margin, not in tomorrow's CPC. SOURCES SYNTHESIZED.

Finding 8 — Sources, scope and method of this synthesis

This analysis cross-reads eight public sector sources published between 2025 and 2026: National Restaurant Association (State of the Restaurant Industry 2025), Lightspeed (Online Ordering Statistics 2025), BrightLocal (Local Consumer Review Survey 2025), Restroworks (Restaurant Social Media Statistics 2025 and Restaurant Coupon Statistics), Capital One Shopping (Gift Card Statistics 2026), Sakari (SMS Marketing Statistics 2025), Rezku (Third-Party Delivery Fees 2026) and CloudKitchens (Delivery App Fees 2026). Not a single figure here is produced in house. SELECTION CRITERION. A source made the cut when the publishing organization is named with an identifiable year, when the metric covers diner behavior or channel cost structure, and when a second source existed to check the direction of the number. The window is 2025-2026; anything earlier was dropped because online ordering behavior shifted too much between 2023 and 2025 to mix the series. HOW SOURCES ARE CONTRASTED. Where two sources disagree, this document shows them side by side instead of averaging them.

Finding 9 — Sources, scope and method of this synthesis — in practice

Delivery commission is the clearest case: Rezku (2026) publishes the nominal 15%-30% range while CloudKitchens (2026) publishes an effective 35%-45% including fees. They are not contradictory, they measure different things, and that distinction is exactly what decides whether your paid advertising toward delivery makes money. HONEST LIMITATIONS. First, most of these sources rest on a US consumer base, so behavioral percentages travel better than absolute dollar values into Latin American or Spanish markets. Second, these are consumer-declared survey answers rather than observed transactions, and stated intent always overstates real conduct; treat them as the ceiling of the range, not the floor. Third, no public source breaks ad spend down by segment with the granularity an owner needs, so the fast casual, full service and QSR split in this analysis is Diego F. Parra's READING of behavioral data, not a published statistical cut. WHAT MASTERESTAURANT CONTRIBUTES.

Finding 10 — Sources, scope and method of this synthesis — key points

The contribution is qualitative and lives in how the data is organized: which figure belongs to which stretch of the sales funnel, which operating decision each one triggers, and what the healthy range looks like by operation size when read against real unit economics. Diego F. Parra signs the interpretation; the figures belong to the cited organizations.