What restaurant marketing strategies work across multiple locations?

The ones that treat the problem as data before creative. Keep hours, menus, prices and attributes correct at every location — that repairs demand you already earned and routinely outperforms campaign work. Manage review velocity per location rather than the lifetime average, because a diner reads the last few reviews and not the mean. Direct promotions at dayparts with idle capacity rather than at the whole week. And run a deliberate programme to convert third-party marketplace customers into direct ones, since every marketplace order builds a relationship you do not own and will have to buy again.

Should a restaurant group use third-party delivery apps?

Almost always yes, and the useful question is a different one. The apps provide reach you cannot replicate, and they take both a commission and the customer record, so the same diner has to be reacquired next week. The strategic decision is what share of demand you are content never to own, and what mechanism moves a marketplace customer to a direct channel — packaging inserts, a first-order incentive on your own ordering path, a loyalty programme worth joining. Groups that never make that decision explicitly tend to discover years later that most of their demand is rented.

How should a multi-unit restaurant measure marketing?

Per location and per daypart rather than per group and per month. Four numbers carry most of it: the share of orders arriving through channels where you keep the customer record, the proportion of locations whose hours and menu are correct today, review velocity per location this week, and incremental covers by daypart. Aggregate covers in particular hide the distinction that decides profitability — between filling an empty Tuesday and adding waits to a Friday that was already full.

Why do templated location pages fail for restaurants?

Because the menu genuinely differs per unit. Prices, availability, regional items and dayparts vary, so a template publishing one menu is wrong somewhere from the first day, and the error surfaces as disappointment at the table rather than as a bounce. The fix is to treat the menu as a data-integration problem: it lives in the point-of-sale, it changes without anyone telling marketing, and until it reaches the page automatically the page will drift out of date at exactly the speed the business changes.

Do restaurant reviews matter more than ratings?

Recency does. A diner choosing where to eat reads the most recent handful, not the average of four hundred, so a location at 4.6 with nothing written in five months can read worse than one at 4.3 with three reviews this week. That makes review velocity per location the thing worth managing, and it has a useful secondary property: a location whose velocity falls to zero is very often the same location with a service problem that nobody escalated.

Restaurant marketing strategies that survive the second location

Most of the revenue on the table is not in a campaign. It is in hours that are correct, a menu that matches the till, reviews written this week rather than last spring, and knowing which dayparts are empty at which locations.

Five things that are true of multi-unit restaurant groups, four numbers worth reporting per location, seven questions for anyone bidding on the work, and three situations where the answer is not to hire.

Published August 23, 2026 · 9-minute read

Five things that are true of a multi-unit group

The delivery marketplace owns the customer, and it is renting them back to you

Every order placed through a third-party app builds that app’s relationship and not yours. You receive the margin after commission and none of the contact record, so the same customer has to be bought again next week.

The strategic question is not whether to be on the marketplaces. It is which share of demand you are willing to never own, and what you are doing to convert marketplace customers into direct ones.

Hours and menu accuracy move more revenue than campaigns

A location showing as closed when it is open, or a menu with last year’s prices, converts nobody. Hours change for holidays, weather and staffing far more often than anyone updates them, and each unit updates them independently or not at all.

This is unglamorous and it routinely outperforms creative work by a wide margin, because it repairs demand you already earned rather than buying more.

Review recency matters more than review average

A diner reads the last few reviews, not the mean of four hundred. A location with 4.6 stars and nothing written in five months reads as declining; one at 4.3 with three reviews this week reads as busy.

Review velocity per location is the metric worth managing. Chasing the average is slow, expensive and mostly invisible to the person choosing where to eat tonight.

Traffic at the wrong hour costs money

A promotion that fills a dining room already at capacity on Friday produces waits, walkouts and worse reviews. The same spend directed at a Tuesday afternoon converts idle labour into revenue.

Marketing that is not aware of the demand curve per location can reduce profit while raising covers, which is how a successful campaign becomes an operations complaint.

The menu is not the same in every location

Prices, availability, local specials and regional items differ per unit, and often per daypart. Templated location pages that publish one menu are wrong somewhere on day one, and the errors surface as disappointment at the table.

The menu is a data-integration problem before it is a content problem. It lives in the point-of-sale and rarely reaches the page automatically.

The four numbers worth reporting per location

Per location and per daypart rather than per group and per month. Aggregate covers conceal the difference between filling an empty Tuesday and adding waits to a Friday that was already full.

Direct versus marketplace mix: The share of orders that arrive through channels where you keep the customer record, tracked per location and over time. This is the number that decides how much of next year’s demand you will have to buy again.
Profile accuracy rate: The proportion of locations whose hours, menu and attributes are correct today. Audit it rather than assume it — the failure mode is silent, and the locations most likely to be wrong are the ones nobody has visited recently.
Review velocity per location: New reviews per location per week, not the lifetime average. Recency is what the diner actually reads, and a location whose velocity has fallen to zero is usually the same one with a service problem nobody escalated.
Incremental covers by daypart: Additional covers attributable to a promotion, split by daypart and location. Aggregate covers hide the difference between filling idle Tuesday capacity and adding waits to a Friday that was already full.

Choosing a restaurant marketing agency

What a firm asks for in the first week tells you what it works on. One that asks for your direct-versus-marketplace mix and your listing accuracy is working on margin. One that asks only for ad account access is working on covers, which is the number most likely to rise while profit does not.

So ask all seven of these, of everyone, including us.

1What share of our orders arrive through channels where we keep the customer record?
2How many of our locations have incorrect hours or menu prices online right now?
3What is review velocity per location this month, and which locations have gone quiet?
4When we promote, do we direct it at dayparts with idle capacity or at the whole week?
5Where does the authoritative menu live, and does it reach every location page automatically?
6What are we doing to convert a marketplace customer into a direct one?
7Which of your deliverables would still be working twelve months after we stopped paying you?

The second is the fastest to check and the most often wrong. Search five of your own locations on a phone tonight and compare the hours shown against the rota.

When you should not hire anyone

While hours and menus are wrong online. Fix the data first; buying demand into an incorrect listing pays to disappoint people.
When the constraint is throughput rather than demand. If Friday is already turning people away, marketing spend belongs in the dayparts that are empty or nowhere.
For a single strong location with a full book. The compounding returns arrive with the second and third unit, where the data stops fitting in one person’s head.

Common questions

What restaurant marketing strategies work across multiple locations?: The ones that treat the problem as data before creative. Keep hours, menus, prices and attributes correct at every location — that repairs demand you already earned and routinely outperforms campaign work. Manage review velocity per location rather than the lifetime average, because a diner reads the last few reviews and not the mean. Direct promotions at dayparts with idle capacity rather than at the whole week. And run a deliberate programme to convert third-party marketplace customers into direct ones, since every marketplace order builds a relationship you do not own and will have to buy again.
Should a restaurant group use third-party delivery apps?: Almost always yes, and the useful question is a different one. The apps provide reach you cannot replicate, and they take both a commission and the customer record, so the same diner has to be reacquired next week. The strategic decision is what share of demand you are content never to own, and what mechanism moves a marketplace customer to a direct channel — packaging inserts, a first-order incentive on your own ordering path, a loyalty programme worth joining. Groups that never make that decision explicitly tend to discover years later that most of their demand is rented.
How should a multi-unit restaurant measure marketing?: Per location and per daypart rather than per group and per month. Four numbers carry most of it: the share of orders arriving through channels where you keep the customer record, the proportion of locations whose hours and menu are correct today, review velocity per location this week, and incremental covers by daypart. Aggregate covers in particular hide the distinction that decides profitability — between filling an empty Tuesday and adding waits to a Friday that was already full.
Why do templated location pages fail for restaurants?: Because the menu genuinely differs per unit. Prices, availability, regional items and dayparts vary, so a template publishing one menu is wrong somewhere from the first day, and the error surfaces as disappointment at the table rather than as a bounce. The fix is to treat the menu as a data-integration problem: it lives in the point-of-sale, it changes without anyone telling marketing, and until it reaches the page automatically the page will drift out of date at exactly the speed the business changes.
Do restaurant reviews matter more than ratings?: Recency does. A diner choosing where to eat reads the most recent handful, not the average of four hundred, so a location at 4.6 with nothing written in five months can read worse than one at 4.3 with three reviews this week. That makes review velocity per location the thing worth managing, and it has a useful secondary property: a location whose velocity falls to zero is very often the same location with a service problem that nobody escalated.

Before you take any of this on faith

The method is visible rather than asserted. howtothink.ai is a public knowledge graph built solo and running in production — 1,700 atomic lessons, roughly 9,700 generated pages, 3,300+ graph edges. It is our own build rather than a client result, which is the point: it is the part you can inspect yourself before anyone asks you for money.

What we actually do for restaurant groups

Scoped to margin rather than covers, because the two move in opposite directions more often than anyone expects.

Hours, menus and attributes correct at every location, audited rather than assumed.
Menus reaching the page from the point-of-sale, so they stop drifting the moment prices change.
Review velocity managed per location, since recency is what a diner actually reads.
Promotions directed at dayparts with idle capacity instead of at the whole week.

How this gets built

The architecture behind it — a canonical record every surface reads from, menus that reach the page from the point-of-sale rather than from a spreadsheet, and review velocity managed per location instead of per brand — is written up in full, with the deployment order and an acceptance test on every phase.

The deployment blueprint Multi-location SEO

Ready to talk instead? Book the 30-minute consultation, or take the three-question diagnostic first. No email required for the diagnostic.

Also: what changes when the units are franchised · keeping hours right everywhere at once