Restaurant local marketing statistics, traced to source

Restaurants are an unusual local vertical: enormous search volume, heavy platform intermediation through delivery aggregators and reservation platforms, and — per our earlier research for this site's strategy — comparatively modest marketing budget per outlet.

The traceable evidence splits into two very different kinds: consumer surveys that are real but either ageing or narrowly scoped, and platform-disclosed financial data that is genuinely audited but answers a scale question, not a discovery question.

Where this site's other statistics pages already carry graded, sourced restaurant-specific data — Google and Meta ad economics, chiefly — this page draws on those rather than duplicating the underlying tables, and links out to them directly. Read against the FOUND, TRUSTED, CHOSEN framework, almost everything traceable here sits at the FOUND stage; the trust and conversion stages are where the published evidence thins out fastest.

How the grading works

The grading is deliberately unforgiving. A statistic from a good organisation still earns a C if the sample size is not published beside the number, because a reader cannot check what they cannot see.

Google is the dominant discovery platform for restaurants — the finding is real, and it's from 2023

A genuine, dated, sampled study, ageing in a category that moves quickly.

What this means for a restaurant's own channel mix

A 2023 finding that Google leads restaurant discovery is still the most defensible starting assumption available, because nothing more recent and comparably sampled has replaced it — not because the number itself is likely to still be exactly 62%. This site's wider local search behaviour statistics cover the same question across local business categories generally, rather than restaurants specifically.

The safer reading is directional: Google almost certainly remains a major restaurant-discovery channel, and the size of its lead over search-specific alternatives like Instagram or a delivery app's own in-app search is the part that has probably moved the most since 2023, in a period that includes a fast expansion of AI-generated search answers and continued growth in delivery-app usage (see below).

A restaurant treating its Google Business Profile as optional on the strength of this ageing study is reading the data backwards — the finding argues for keeping the highest-confidence channel current, not for waiting on a fresher number before investing in it — which our Google Business Profile optimisation guide covers step by step.

The newest dated study on restaurant choice measures travellers specifically, and personal recommendation still beats every channel it tested

The National Restaurant Association's 2026 Summer Travel and Dining report is the freshest sampled consumer study this page identified — real, dated to May 2026, and genuinely new to this page. It measures a narrower population than 'diners' generally: people taking a domestic trip this summer, choosing where to eat away from home.

What this means for a restaurant that depends on tourist or visitor trade

These four figures describe the same underlying decision from different angles: a traveller is unusually likely to try somewhere new (88%), unusually likely to lean on a recommendation rather than research it themselves (87%), and still price- and location-sensitive on top of that (90% and 70%).

For a restaurant in a location that draws visitors, the practical implication is that being recommendable — by staff at a hotel front desk, by a local's answer to "where should we eat", by a review a stranger reads once and repeats — plausibly does more work than any single digital channel this page has evidence for.

None of that is measurable the way a Google click or a delivery-app order is measurable, which is exactly why it tends to get under-invested against channels that produce a dashboard.

This finding doesn't replace the 2023 Google-discovery data above — it measures a different population (travellers, not diners generally) making a different kind of decision (unfamiliar location, no existing habit) — but the two sit compatibly together: word-of-mouth for the initial nomination, then very plausibly a Google search or a check of the restaurant's own Google Business Profile to confirm hours, location and menu before actually going, a sequence this survey wasn't designed to measure directly.

Word-of-mouth and online reviews are related but not identical trust signals — this site's own local reviews statistics page covers how much weight the written, public version of a recommendation carries once a stranger, not a friend, is doing the recommending, and our local review strategy guide covers how those reviews get earned deliberately rather than by chance.

Delivery and reservation platforms disclose real, audited scale — and neither DoorDash, Uber nor OpenTable's parent discloses how a diner actually finds a new restaurant on them

DoorDash and Uber are public companies; both file audited financial disclosures with the SEC every quarter, the same source type this site used for its Yelp statistics page. Those filings are a genuinely stronger source than most of what circulates about aggregator platforms — and they answer a narrower question than 'how does restaurant discovery work' would suggest.

Platform	Metric	Latest figure	YoY change
DoorDash	Marketplace monthly active users (FY2025 year-end)	56 million+	Not disclosed as a YoY %
DoorDash	Marketplace Gross Order Value (Q4 2025)	$29.7 billion	+39%
DoorDash	Marketplace orders (Q4 2025)	903 million	+32%
Uber	Delivery segment Gross Bookings (Q2 2026)	$27.46 billion	+26% (constant currency)
Uber	Total Monthly Active Platform Consumers, all segments (Q2 2026)	208 million	+16%

Figures are not directly comparable to each other: they cover different reporting periods, different business scopes (DoorDash's figures include grocery and retail; Uber's Delivery figure includes non-restaurant categories and Uber's MAPC figure includes ride-hailing and freight), and different metric definitions. They are placed together here to show disclosed order of magnitude, not to rank the two platforms. Source: DoorDash, Inc. and Uber Technologies, Inc., SEC filings

Neither company's public disclosures were found, in the sources reviewed on 8 September 2026, to break out a restaurant-specific active-user count, a new-restaurant-discovery rate, or any measure of how often a customer opens the app already knowing which restaurant they want versus how often they browse and pick one they haven't ordered from before.

Booking Holdings, OpenTable's parent company, discloses even less at the OpenTable level specifically: its Form 10-K for fiscal year 2025 names OpenTable as "a leading brand for booking online restaurant reservations" and groups its revenue with advertising, without publishing a restaurant-network count, a seated-diner figure, or a reservation volume anywhere in the filing.

Grubhub, meanwhile, was sold out of Just Eat Takeaway in 2024 and is now owned by the privately held Wonder Group, which files no comparable public disclosures at all — meaning even the scale-only data available for DoorDash and Uber has no equivalent for Grubhub.

What this means for weighing aggregator exposure against direct discovery

Scale and discovery are different claims, and the aggregators' own audited numbers only support the first one. A restaurant can conclude, correctly, from the figures above that these platforms move tens of billions of dollars a quarter and reach tens of millions of monthly users.

What those figures give no basis for, from any of these companies' own disclosures, is a conclusion about what share of that reach translates into a diner finding that specific restaurant for the first time, as opposed to reordering from a restaurant they already use the app to visit.

That gap is exactly what this page's original 2026 review flagged as a research gap on delivery-aggregator influence. Closing it took confirming, with primary sources this thorough, that the gap is real rather than a case of this page simply not having looked hard enough — the standard our research methodology sets before a gap can be claimed.

Where restaurant ad economics have already been measured on this site, restaurants convert unusually well on both Google and Meta

This site's local lead generation and Meta Ads statistics pages already carry graded, sourced rows for restaurants inside larger industry benchmark tables. Rather than duplicate those records under a second URL, this section draws the comparison those two tables make possible once restaurants are read against every other vertical in them.

LocaliQ's search-advertising benchmark table records Restaurants & Food at a $2.05 average cost per click, an 8.05% click-to-lead conversion rate and a $30.57 cost per lead on Google and Microsoft Ads search campaigns — see the full 23-industry table and its grading on this site's local lead generation statistics page.

WordStream's Meta Ads benchmark table, covering Facebook and Instagram Leads-objective campaigns, records Restaurants & Food at a $0.74 average cost per click, an 18.25% conversion rate and a $3.16 cost per lead — the cheapest of the eight local-service categories in that table, detailed on this site's Meta Ads statistics page.

Read against those tables' own all-industry averages — $66.69 for search, $27.66 for Meta — restaurants convert an ad click into a lead at a markedly higher rate than most other local verticals measured on both platforms, and the resulting cost per lead lands well below almost every other category in either table.

Neither report explains why, and this page won't invent a reason where the source doesn't supply one.

A plausible interpretation, clearly labelled as interpretation rather than measurement, is that the action being counted as a "lead" — a call, a directions request, an online order — asks far less commitment of the customer than a home-improvement quote request or a legal consultation booking, which would lower the measured cost of the action without saying anything about whether the resulting customer is worth more or less than one from a higher-friction category.

What this means for budgeting restaurant ad spend

A restaurant benchmarking its own Google or Meta cost per lead against either table's blended all-industry average is comparing itself to the wrong number twice over — once for the general reason this site's lead generation page already makes (the blended average obscures a near-5x industry spread), and a second time because restaurants specifically sit well below even their own category peers on both channels.

A restaurant paying close to $30 per lead on search or $3 on Meta is roughly in line with what the category actually costs. A restaurant paying several multiples of either figure has more room to investigate the campaign, the offer or the landing experience than a blended-average comparison would suggest — a judgement that gets easier once the value of a single lead has been worked out rather than assumed.

Customers say they want app-based ordering — but the industry's own figures aren't precise enough to budget against

The National Restaurant Association's Restaurant Technology Landscape Report 2024 is the most recent NRA study this page identified that touches on how customers say they want to interact with restaurant technology — real, named, and reported in rounded fractions rather than precise percentages on the page this review could access.

What this means for prioritising a restaurant's own ordering technology

Three rounded fractions from a single named, dated source aren't enough to plan a technology budget against precisely, and this page grades them C for exactly that reason: real organisation, plausible method, no published sample size on the page available without payment.

What they're strong enough to support is direction, not precision — customers report wanting app-based ordering more than a "some customers prefer it" hedge would suggest, and deal-seeking behaviour (7 in 10) is common enough that a restaurant with no visible offer is very plausibly leaving interested, price-sensitive customers with nothing to act on — the kind of on-site gap our local website conversion guide works through.

Where this page needs a number precise to the decimal point — setting an actual marketing budget, for instance — the ad-economics section above, sourced to disclosed platform performance data rather than a survey summary, is the stronger source to build on.

What the evidence covers

Most of the consumer-survey data on this page is US-specific. The National Restaurant Association's 2026 Summer Travel and Dining report and the Restaurant Technology Landscape Report 2024 both sample US consumers only; PYMNTS' 2023 Connected Dining data is US-only too.

The platform-disclosed figures (DoorDash, Uber) are global in scope but not broken out by country in the filings reviewed. No equivalent, comparably dated restaurant-discovery research for markets outside the US — including India — was identified in the sources reviewed on 8 September 2026.

The numbers that did not survive the trace

These appear in current statistics roundups. Each one fails on provenance, and quoting them transfers the failure to you.

How this page is made and how to cite it

The PYMNTS, National Restaurant Association and platform-filing figures were each traced to their original source directly — the SEC filings and 8-K exhibits linked in the records above, not a secondary summary of them.

Last verified: 8 September 2026. Reviewed quarterly — this category deserves a fresher, broader consumer-discovery study than any single source on this page currently provides, and one may exist that wasn't surfaced in this review.

Preferred citation Parashar, S. (2026). Restaurant local marketing statistics, traced to source. shrutiparashar.com/statistics/restaurant-marketing/

Questions

How do most people find a restaurant to try?

Google was the dominant platform in the most recent traceable study — 62% of consumers used it to search for restaurant information, more than any other single digital platform, per PYMNTS' Connected Dining Report (n=2,200+ US consumers, May 2023). That data is three years old in a category where delivery-app and social-discovery behaviour changes quickly, so treat it as directional rather than current. A separate, fresher 2026 study of travellers specifically found personal recommendation from family and friends (87%) outranking every channel it tested when choosing an unfamiliar restaurant — a different population making a different kind of decision, not a contradiction of the Google finding.

How much do delivery aggregators like DoorDash or Uber Eats affect restaurant discovery?

Both companies' SEC filings disclose real scale — DoorDash reported 56 million-plus monthly active Marketplace users at the end of 2025 and $29.7 billion in Q4 2025 Marketplace Gross Order Value; Uber reported $27.46 billion in Q2 2026 Delivery segment Gross Bookings. Neither filing breaks out how many of those users discovered a restaurant they hadn't ordered from before, as distinct from reordering from one they already use the app to visit. No substantial, dated, sampled study specifically measuring aggregator platforms' share of new-restaurant discovery — as distinct from ordering volume — was identified in the sources reviewed on 8 September 2026. That's a genuine, notable gap given how central these platforms have become to the category.

Do DoorDash and Uber disclose how many restaurants are on their platforms?

Not in the filings this page reviewed. DoorDash's Form 10-K for fiscal year 2025 and its Q4 2025 earnings release disclose Marketplace-wide user, order and Gross Order Value figures but no restaurant-specific merchant count — the disclosed figures blend restaurant, grocery and retail categories together. Uber's Q2 2026 earnings release discloses Delivery segment Gross Bookings but not a restaurant-only breakout, and its total Monthly Active Platform Consumers figure spans ride-hailing, delivery and freight combined, not delivery alone.

Does OpenTable publish how many diners it seats or how many restaurants use it?

Not at a level this page could verify. OpenTable's parent company, Booking Holdings, names OpenTable in its Form 10-K for fiscal year 2025 as 'a leading brand for booking online restaurant reservations' and groups its revenue together with advertising, without publishing a restaurant-network size, a seated-diner count, or a reservation-volume figure anywhere in the filing. This is a thinner disclosure than either DoorDash's or Uber's, both of which at least publish segment-level scale figures.

What does the newest research actually say about why travellers pick a restaurant?

The National Restaurant Association's 2026 Summer Travel and Dining report (May 2026, 1,000+ US consumers) found 88% of surveyed domestic summer travellers plan to visit a restaurant they've never tried before on their trip, 87% rely on recommendations from family and friends when choosing an unfamiliar restaurant, and 90% cited convenient location and affordability as important factors. This measures travellers specifically, not diners generally, and the report doesn't test digital discovery channels against word-of-mouth in the same question — so it's evidence that recommendation matters enormously to this population, not proof it outranks a Google search or online reviews specifically.

Are restaurants cheaper or more expensive to advertise for than other local businesses?

Cheaper, on the evidence this site has gathered for both major ad channels. LocaliQ's search-advertising benchmark table records Restaurants & Food at a $30.57 cost per lead against a $66.69 all-industry average; WordStream's Meta Ads benchmark records Restaurants & Food at $3.16 per lead against a $27.66 all-industry average — the cheapest of the eight local-service categories in that table. Both figures live in full, with grading, on this site's local lead generation statistics and Meta Ads statistics pages respectively; this page draws the comparison rather than duplicating the underlying tables.

Is it true that adding photos to a Google Business Profile gets a restaurant 42% more direction requests?

That specific figure — 42% more requests for driving directions, 35% more website click-throughs — could not be traced to any current, live Google source. Google's own Business Profile Help documentation on photos, checked 8 September 2026, makes no comparison claim of this kind. The number circulates widely across marketing content, almost always attributed only to 'Google' with no working link, which is the pattern this site treats as a do-not-cite statistic rather than a real, checkable one.

What do restaurant customers say about ordering technology?

The National Restaurant Association's Restaurant Technology Landscape Report 2024 found roughly 8 in 10 delivery customers and 7 in 10 limited-service customers say they'd be likely to order using a smartphone app, and roughly 7 in 10 adults look for deals across takeout, delivery and dine-in. These are rounded fractions from a named, dated 2024 source, not precise percentages — the publicly accessible summary doesn't disclose an exact consumer sample size, so this page grades the figures C: real and plausible, not independently auditable at the number available.

Is Grubhub's scale disclosed the same way as DoorDash's or Uber's?

No. Grubhub was sold out of Just Eat Takeaway in 2024 and is now owned by the privately held Wonder Group, which doesn't file the quarterly and annual disclosures a public company does. That means even the scale-only figures available for DoorDash and Uber — monthly active users, Gross Order Value, Gross Bookings — have no current public equivalent for Grubhub specifically.

What's the single biggest gap in restaurant local-marketing research right now?

The gap between disclosed platform scale and disclosed discovery behaviour. DoorDash and Uber both publish real, audited figures about how much money and how many users move through their platforms every quarter — and neither discloses how a diner actually finds a specific, new-to-them restaurant on the app, as opposed to reordering from one they already use it for. That's the same gap this page's original review flagged in 2026, confirmed here with primary-source depth rather than resolved — no substantial study closing it was identified in the sources reviewed on 8 September 2026.

Related reading: our research methodology , Local search behaviour statistics , Google Business Profile statistics , local reviews statistics .

The neighbouring research sits in the statistics archive, 275 graded figures across 24 pages.