Paid advertising for restaurants: the 2026 numbers and the decision each one triggers

GROUP 1 · Local discovery: where your demand actually lives

Seventy-six percent of people searching for something nearby on a phone walk into a physical business within the next 24 hours, per BrightLocal (Local SEO Statistics 2026), and that share climbs to 88% when the same report measures all local mobile searches. Add that «food near me» queries grew 99% year over year according to Restroworks (Google Restaurant Search Statistics 2024), and that more than 60% of restaurant searches start on a handset, and the map draws itself: your demand does not live in a social feed, it lives on a map with a radius sketched around your empty table. The decision these four numbers trigger together annoys anyone who already signed with a content agency, because it is not about raising the ad budget but about completing the Google Business Profile listing and confirming that hours, menu and cover photo belong to 2026. Local discovery is your own asset; paid media only amplifies what that listing already holds up.

The free click your listing gives away before you pay for one

Forty-two out of every hundred local searchers click inside Google's map pack, per Semrush 2025 as reported by Malou (Local SEO for Restaurants), and 79% of restaurant searches are NON-BRAND in that same study: nobody types your name, they type «pizza near me» at 8:40 p.m. while hungry. There is a tension worth resolving before you sign any budget: the same owner who argues over a 1.20-dollar cost per click hands away 42% of map-pack traffic without blinking, keeping pandemic hours and three photos from 2021. When the listing is incomplete, every advertising dollar pushes people toward a shuttered storefront. Correct sequence is listing first, reviews second, paid media last, and reordering it is the most expensive way to learn local SEO. Advertising within 3 to 5 kilometres of the door instead of across the whole city is the one targeting decision that moves the P&L, and the reason is arithmetic rather than ideological.

GROUP 2 · What geotargeted advertising costs against city-wide advertising

With 76% of nearby searches ending in a visit inside 24 hours (BrightLocal 2026) and 42% of local clicks concentrated in the map pack (Semrush 2025 via Malou), every impression served to someone forty minutes away by car buys intent your kitchen will never collect. Back to the Bogotá case from March 2026: 4,180 dollars spent, 1.9 million impressions, 812 clicks. Eight hundred twelve clicks against 1.9 million impressions works out to a 0.043% CTR, and real cost per click lands at 5.15 dollars. None of those three figures appeared in the agency report. CPM and reach did, and those are awareness metrics sold as sales metrics. Take the same spend, cut the radius to five kilometres and accept losing 80% of your reach: impressions drop from 1.9 million to roughly 380,000, yet the audience left is the one that can walk over or drive fifteen minutes.

What would happen if you moved those 4,180 dollars to the square metre?

If CTR rises from 0.043% to a reasonable 0.35% on a warm audience, you get 1,330 clicks for the same 4,180 dollars, meaning 3.14 dollars per click against 5.15.

And if 8% of those 1,330 clicks turn into a booking or an order at a 22-dollar average check, the campaign returns 2,340 dollars in direct sales plus the repeat table. That calculation needs no expensive tool, it needs someone to write it down. What I see again and again is that nobody writes it, because the agency report arrives looking handsome and the till does not complain until the following quarter. A paid click landing on a badly built menu is money turned into curiosity.

GROUP 3 · The conversion that happens after the click, inside the menu

Trade figures are blunt here: a complete digital offer —browsable menu, ordering and payment in one flow— lifts the check between 20% and 30% per Sunday (QR Code Ordering 2025), QR-code ordering raises check size by 9% versus traditional table service in that same report, and menu-psychology techniques add 15% or more without touching a single price, per NeatMenu (Menu Psychology 2026). Stacked together, those three levers move average check far beyond any saving you might squeeze out of cost per click. The mini-conclusion of this group stings anyone who only watches the ads dashboard: before optimising the campaign by 10%, fix where the click lands, where 20 to 30 points are sitting idle. Destination first, traffic afterwards. Buying the same guest twice costs a fraction of buying them the first time. Paytronix (Effectiveness of Loyalty Programs 2025) measures a 16.5% year-over-year rise in member spend when targeting runs one to one, and Technomic 2026, cited by Restroworks, records a 19% annual increase in limited-time offers across the sector.

GROUP 4 · Repeat business: promotions, loyalty and the spend of whoever already came

QR scan volume grew 433% in two years per QR Code (Restaurant Usage 2025), which means the channel for capturing that data already sits on the table, literally. I got this wrong for years, recommending paid acquisition first and the loyalty programme afterwards, when the profitable order runs the other way. The decision these figures trigger: assign 30% of your ad budget to audiences built from your own database, and watch cost per sale drop without negotiating a cent with the platform. Diego F. Parra, restaurant consultant and founder of Masterestaurant, judges every paid campaign with one extra column added to the report: how many of those clicks ended up seated. It sounds obvious and almost nobody does it, because it demands tying the ad to the booking, the booking to the check and the check to contribution margin, and that knot is operations work, not agency work.

Why Masterestaurant measures advertising against the till, not against the dashboard?

Under the Masterestaurant framework, paid media enters the break-even calculation as a fixed monthly cost, never spread across plates, and it carries a clear ceiling:

if the campaign fails to return three times its cost in attributable sales within 30 days, it goes dark and the money moves to the listing, the reviews and the digital menu. That 3x threshold is not a whim, it is the minimum a 30% food cost survives with payroll and rent already committed. Seventy-six percent (BrightLocal 2026): that share of nearby mobile searches ends in a physical visit within 24 hours. Action: block two hours this week, complete your Google Business Profile with 2026 hours, menu and photos, and ask five guests from this week for a review. Seventy-nine percent (Malou 2025): restaurant searches that are non-brand. Action: build your campaigns and your copy around the category and the neighbourhood —«breakfast in Chapinero»— rather than around your own name, which hardly anyone types.

The 3 numbers you should tattoo on yourself

Twenty to thirty percent (Sunday 2025): what a complete digital offer with menu, ordering and payment adds to the check. Action: run the flow from your own phone, time how many taps sit between the ad and the payment, then delete two. If only one of the three fits this week, do the first one today. GROUP 1 · Local discovery. Google reports that 76% of people searching for something nearby on a phone visit a physical business within 24 hours, and 46% of all searches carry local intent. In cash terms: most of your demand is not sitting on a social feed, it sits on a map. The decision this triggers annoys anyone who just signed a content agency — before raising the social budget, finish the Google Business Profile and confirm that hours, menu and cover photo belong to 2026. Group takeaway: local discovery is an owned asset, and paid media only amplifies what the listing already holds up.

The numbers grouped: where restaurant paid advertising is actually decided

GROUP 2 · What geotargeted media costs. Average cost per click in the restaurant vertical sits near 1.95 USD on search, while CPM on short-radius social targeting runs 8 to 14 USD, per WordStream's 2025-2026 benchmarks. Alone the number says nothing; the cross with conversion decides. At 1.95 USD per click and 9% conversion to a booking, a new guest costs 21.60 USD, and with a 26 USD check at 57% margin you keep 14.80: the campaign survives on fumes. Takeaway: without lifting landing or listing conversion, chasing a cheaper CPC is cosmetic. GROUP 3 · Delivery and commissions. Platforms charge between 15% and 30%, with the top band near 27-30% for bundles that include visibility. Advertising a 34% food cost dish inside the app produces negative margin before packaging enters the calculation. This decision belongs to the menu, not to marketing: build a separate delivery menu, dishes under 28% food cost, priced to absorb the commission.

The numbers grouped: where restaurant paid advertising is actually decided — in practice

Group takeaway: in delivery, paid advertising turns profitable only after menu engineering, never before it. GROUP 4 · Reviews and trust. Some 93% of consumers say online reviews shape their buying decisions, and half a star of rating difference moves booking volume measurably. A 3.9★ listing pays more per click because it converts worse, so the effect compounds: weaker rating, weaker conversion, higher acquisition cost. The decision fits in one line — make the review request a shift-closing task with a named owner. Takeaway: 5★ reviews are the cheapest discount available on your media invoice. GROUP 5 · Guest LTV and time horizon. A recurring neighbourhood guest averages four to five visits a year, which turns a 26 USD check into 112 USD annually and completely changes what you can pay to acquire them. Whoever measures at thirty days switches off campaigns that were working. The decision: cap acquisition cost at 25% of twelve-month LTV, not at first-visit margin.

The numbers grouped: where restaurant paid advertising is actually decided — key points

Group takeaway: gastronomic growth is won on the second order, and that one is bought with product and recall, not with more impressions. THE 3 NUMBERS WORTH TATTOOING. First: 76% of local searches end in a visit within 24 hours — action, finish your Google Business Profile this week and publish eight new photos. Second: 27% average delivery commission — action, pull every dish above 28% food cost off the delivery menu and reprice it today. Third: 4.3 annual visits per recurring guest — action, calculate your real LTV and cap acquisition at 25% of it before approving the next media budget.