More people decide on your restaurant in Google Search and Maps than on your website. They check hours, photos, the menu, and whether the place looks worth the trip. Run that profile as a live storefront: categories, hours, photos, posts, and review replies that match the door. Then report each location on its own line in calls, directions, website reservations, nearby foot traffic, and confirmed visits where available. Hire someone who will tell you when a location is slipping, and what a number cannot prove.
Guests decide on Search and Maps before they reach a website. They compare hours, photos, reviews, the menu, and whether the place looks worth the trip. A complete, current Google Business Profile is the surface that decision happens on. Local SEO is the discipline of keeping that surface accurate, visible in the right trade area, and measured in the actions that fill a room: calls, direction requests, website reservations, nearby foot traffic, and confirmed visits where Google can show them.
The work is unglamorous and specific. Categories, hours, attributes, and the menu have to match reality. Photos and posts have to stay fresh. Review replies have to happen. For a group, each address is its own storefront. A blended account that averages three rooms together will hide the one that is slipping.
Key takeaway: Maps is where nearby guests choose. The website is where some of them book. Treat the profile as the front door, not a listing you set once.
You do not need a media buyer to see whether the storefront is honest. Open the profile the way a guest would, on Maps, for every address.
| Categories | The primary category should match how a nearby guest would search the room: restaurant, steak house, sushi restaurant, or the actual type. Extra categories should be true of this address, not a wish list from another concept. |
|---|---|
| Hours | Service hours, kitchen hours, and holiday hours have to match the door. A wrong close time is a guest who arrives to a dark dining room and a one-star review. |
| Name, address, phone | The name guests will type, the real address, and a number that a person answers. If the phone rings a call center that cannot book the room, the profile is sending demand into a hole. |
| Photos | Current food, room, and exterior photos. A guest deciding tonight is comparing rooms. Dated or stock photos lose that comparison. |
| Posts | A recent update for hours, a seasonal menu, or a private-dining note. An empty post history looks like an abandoned listing. |
| Review replies | Reply to 4- and 5-star reviews with something specific. Ignore 1- to 3-star reviews for public reply. Silence on the good reviews makes the profile look unmanned. |
| Menu and booking path | The menu, reservation link, and website should go to this location. A group that points every pin at one homepage is asking nearby guests to guess. |
If those rows fail, more ads will not make the restaurant easier to choose. Fix the storefront first. Then look at whether nearby guests can find you at all.
Key takeaway: Ownership starts with an owner check. Categories, hours, photos, posts, and replies are weekly work, not a setup project.
Run the profile as a live channel. Completeness, photos, posts, menus, and review replies are maintenance, not a one-time setup. The measure is profile actions by location, not a vanity view count with no follow-through.
Win the searches a restaurant can win. Head terms like "best restaurants" belong to aggregators and food press. The winnable ground is the map pack, the business profile, and specific long-tail pages: private dining, a signature dish, a neighborhood, an occasion.
Keep a per-location scoreboard. Each restaurant is measured against itself, month over month, with its own trade area and its own search terms. Brand demand is separated from genuinely new demand so the spend does not flatter itself.
Report the actions that make money. Views are context. Calls, directions, website reservation clicks, nearby foot traffic, and confirmed visits where available are the score. When a term or a location is not worth the money, that call is made in writing.
Key takeaway: Local SEO for a restaurant is Maps findability plus an honest storefront. It is not a blog calendar and it is not a citywide "best of" campaign.
The dining-room scoreboard on Maps is useful only if you know where each number stops. Do not treat every tap as a seated guest.
| Profile views | Someone saw the listing. That is context. It does not prove they called, asked for directions, or booked. |
|---|---|
| Calls | Someone tapped the number. Useful. It is not a website reservation and it is not a confirmed visit. |
| Directions | Someone asked how to get there. For dinner, that is a strong nearby-guest signal. It is still not proof they sat down. |
| Website reservation clicks | Someone left Maps for your booking path. Count that with the reservation system, not as a seated cover by itself. |
| Confirmed visits | Use this when Google can show it for the location. Say so when it cannot. Never invent a visit count, and never borrow another restaurant's visits. |
| Bookings on your own site | Hotel and restaurant bookings that land on the property's own site belong on their own line. A Maps click is not that booking until the booking system says so. |
Private-event inquiries stay off this scoreboard unless someone actually sent the events form. A directions tap from a planner is still a map tap, not a booked buyout. The FAQ keeps dining-room actions and sent event inquiries on two scoreboards for that reason.
Translation: Report calls, directions, and confirmed visits by location. Label what each number cannot prove. Do not average the rooms.
Hire the partner who will show you each location on its own line, name the proof, and tell you when to spend less. Ask for the actual search terms the budget bought. Ask who owns the ad accounts and the Business Profile login. Ask what one confirmed visit or one website reservation is worth, and whether the monthly report will show that number.
A one-time listing setup is not the job. A blended dashboard is not the job. Press is a different hire. If nearby guests cannot find the restaurant, start with the storefront checks above, then use the seven-question vetting guide if you are comparing shops.
Hire this for same-metro groups, Peninsula groups, and groups expanding into a new city. Do not hire this if you want one blended dashboard, a one-time listing setup, or press.
Key takeaway: The profile is a live storefront. If the report cannot show calls, directions, and confirmed visits by location, it is not doing the job.
Kirk's Steakburgers is the same-metro example, not a multi-city one. Campbell, Palo Alto, and Santa Clara each need to show up when nearby diners open Maps, with each room readable on its own line. Hours, photos, and review replies cannot be copied from the busiest room and pasted onto the quieter one.
Nisei is the destination-restaurant example. For a tasting-menu room, the Google profile is often the storefront guests actually use: hours, photos, directions, and the path to a reservation. The monthly report should show those actions, not a blended view count.
Gott's is named here for the pattern, not for a result. Eight Northern California restaurants, the Los Angeles Farmers Market location already open, and Santa Monica coming soon. The job is to keep each new trade area honest as the brand grows, without treating expansion as one blended coastal average, and without borrowing another restaurant's visit count.
Key takeaway: Compare each address with itself. A group average hides the room that nearby guests stopped finding.
The profile is the front door, and it needs weekly attention, not a one-time setup. Each address gets its own scoreboard. "Best restaurants in the city" belongs to the review sites; the map pack, your neighborhood, and your signature dish are winnable.
My advice: Ask for calls, directions, and confirmed visits by location, not a views count. If that line is missing, the report is decorating the storefront instead of running it.
Look for a partner who reports each location against itself on the profile and on Maps. Jay Street Media does this work. Kirk's Steakburgers is the same-metro example: Campbell, Palo Alto, and Santa Clara each get their own line. Gott's Roadside is named for the expansion pattern, not for a visit result: eight Northern California restaurants, the Los Angeles Farmers Market location open, and Santa Monica coming soon.
Keeping the profile complete and current: categories, hours, menu, photos, posts, and review replies, then measuring the actions that follow (views, calls, and direction requests) by location. For a group, each address is its own storefront.
Someone who will not blend the rooms into one average. Each location has its own trade area and its own scoreboard. The Kirk's work is the same-metro version of that. The Gott's work is the expansion version.
Confirm the primary category matches how guests search the room, hours and holiday hours match the door, the menu and photos are current, a recent post exists, and 4- and 5-star reviews have a reply. Then look at calls, directions, and website reservation clicks by location, not a blended view count.
Usually no, not action by action. A call is not a website reservation. A directions tap is not a confirmed visit. A confirmed visit, where Google can show it, is still not a named guest in your reservation book. Use those actions as the dining-room scoreboard, say where each number stops, and do not treat profile views as proof that nearby guests found you.
Published by Jay Street Media, a San Francisco team that does this work for restaurants and hotels. The guide stands on its own, and it tells you plainly which parts you can do yourself.