Restaurant photos that actually affect your ranking

Most restaurants treat photos as decoration. They are closer to signage. A photo set is the only part of your listing that shows a stranger what happens to them if they walk in, and it is doing that job whether or not you chose the images.

The uncomfortable part is that if you stop uploading, customers keep uploading for you. Their photos are lit by a phone flash at nine at night and they will sit at the top of your profile indefinitely.

Do photos actually change where I rank?

Not directly, and anyone who tells you otherwise is selling a photo package.

What photos change is what happens after your listing appears. Somebody sees three restaurants in a map result and picks one. A profile with a bright exterior, a readable dining room, and food that looks like the food gets picked more often than a profile with one dark photo from 2019.

That choice is not invisible to Google. In Whitespark's 2026 Local Search Ranking Factors survey, where 47 local search experts scored 187 factors, click-through rate from the local pack and Maps results ranks thirteenth. Completeness of the profile ranks sixteenth. Neither is a photo factor as such, and both are downstream of whether your listing looks like a place worth tapping.

Grade the survey honestly. Whitespark sells local SEO software and the report is expert opinion, not controlled testing. It is still the most careful public attempt at ordering these signals.

There is a second effect that is easier to miss. Fresh photos are one of the signals that a business is currently operating. A profile whose last upload was two years ago reads like a place that might be closed, which is the same problem a profile with no recent reviews has.

So the honest version is this. Photos are not a ranking factor, they are a conversion factor that eventually behaves like a ranking factor.

Do keywords in filenames and alt text help my listing?

They do essentially nothing for local ranking, and this is where most restaurant photo advice quietly wastes an afternoon.

The same Whitespark survey publishes its lowest-scoring factors, meaning the things the panel believes do almost nothing. Sitting near the bottom of 187: presence of keywords in image metadata fields, keywords in image filenames, and keywords in image ALT attributes. All three score in the low forties or below on a scale where the top factor scores 227.

That does not make alt text worthless. It is an accessibility requirement, it helps a screen reader, and it belongs on your website for that reason alone. It is simply not a lever on your map position, and renaming 200 files to `korean-bbq-san-jose-brisket-12.jpg` is not a marketing project.

There is a related instruction that has become genuinely risky rather than merely useless. Whitespark's 2026 edition added new negative and suspension-risk factors, and two of them are about media: presence of AI-generated photos or videos on the profile, and presence of AI-generated text content on the profile. A generated hero image of a dining room you do not have is no longer a shortcut. It is on the list of things that can cost you the listing.

Shoot the room you actually have. Name the file whatever you want.

Which photos does Google actually want?

There is a short list, and most restaurants are missing at least two items on it.

The exterior, in daylight, from the direction people approach. This is the single most useful photo you own and almost nobody has a good one. It answers the only question a nervous first time customer has, which is whether they will recognize the place when they get there. If your restaurant sits in a strip center or behind a shared entrance, this photo prevents a bad review before it happens.

The interior, lit the way it is actually lit. One wide shot of the dining room during service. Not empty at ten in the morning. A room with people in it reads as a room worth being in.

Food, one dish per frame, shot from above or at plate level. Six clean dish photos beat forty. Pick the six you actually want to sell, not the six that photograph best.

The menu, as photos, in addition to a real menu page. People screenshot menus from profiles constantly. This does not replace a text menu on your site, it supplements it.

One photo of the team. It does more for a local business than anyone expects, and it is the one photo an aggregator can never have.

Everything else is optional. Bathrooms, close ups of garnish, and stock photography of generic food are noise.

How often should I be adding new photos?

Often enough that the top of your profile is never a year old.

A monthly cadence is enough for most restaurants. New special, new season, a full room on a Friday, a new hire behind the pass. It takes a manager four minutes and it is the kind of thing that should sit on a recurring list rather than in someone's memory, alongside the rest of what an owner should check on their listing every month.

The failure pattern is a burst. An owner hires a photographer, uploads forty images in one afternoon, and never uploads again. Two years later the profile looks abandoned even though the food improved.

Steady beats large. That is true of photos, and it is true of most things that feed a local listing.

Do I need to hire a photographer?

For the food, probably once. For everything else, no.

A phone in daylight near a window will produce a better dish photo than a mediocre professional shoot under warm restaurant lights. Turn the overhead lights off, use the window, shoot from directly above or straight across, wipe the plate rim.

Where a professional earns their fee is the room and the exterior, because those need equipment and patience you do not have during service. Get those done once and reuse them.

What does a photo shoot have to return to be worth it?

Here is the arithmetic. Replace every input with your own.

Say a half-day shoot for the exterior and the dining room costs you $900.

Pull your average check per person from your POS. There is no credible published US average check for independent restaurants, so a national figure would be worse than the one already in your system. Say yours is $38.

The National Restaurant Association's 2025 Operations Data Abstract, built from more than 900 operators, puts food and non-alcohol beverage cost for full-service restaurants under $2 million in annual sales at a median of 33.7% of sales in 2024. So each additional cover contributes roughly $38 less 34 cents on the dollar, or about $25.19, before any extra labor those covers require.

• $900 divided by $25.19 is about 36 covers.
• 36 covers is roughly one extra four-top a week for two months.

That is the whole hurdle. If a better exterior and a real interior shot cannot produce one additional four-top a week for two months, the shoot was the wrong spend. Most operators, once they see the number, conclude it is worth doing once and not worth doing annually.

Run it with your own numbers. If your check average is $18 rather than $38, the hurdle roughly doubles, and phone photography by the window becomes the correct answer.

Where does this advice break down?

In three places.

When you are losing on distance, not on presentation. Proximity of your address to the searcher is the second heaviest factor in the Whitespark survey. If a better restaurant four blocks closer is taking the result, a photo set does not move you above them at that spot, and how to rank when you are not the closest restaurant is the actual problem.

When the photo oversells the plate. If the image shows a portion larger than what leaves your kitchen, you have bought yourself a complaint and probably a comp. I learned to read a plate cost line before I learned to read a marketing report, and the discipline is the same one behind reading a restaurant P&L properly: the number and the plate have to match what is actually happening in the building.

When the profile is not the surface being read. Google's own documentation says a Business Profile helps a local business appear "in both AI responses and other Google Search results," and the same record feeds both. But when a summarizer builds an answer about restaurants nearby, it is working from ratings, review text, hours and prices rather than from your dish photography. Photos still matter there, because they decide whether the person who reads the summary taps through. They are not what got you into the summary, which is a separate discipline covered in how a restaurant ends up inside AI recommendations.

What to do this week

Open your profile on a phone and look at the first three photos a customer sees. If any of them were uploaded by a customer, you have a gap to fill.

Take the exterior shot today, in daylight, standing where a customer parks.

Shoot six dishes by a window. Ten minutes total. Upload them.

Put a photo upload on a monthly recurring task with a name attached to it, not a department.

Run the payback arithmetic above with your own check average before you book anybody. The number decides the scope of the shoot.

Then check whether your food photos match what you actually want to sell, because the searches you show up for and the dishes people order are connected. If your goal is larger orders rather than singles, the photo set changes entirely, and that difference shows up in how restaurants get found for catering searches.

Questions about your photo set?

Email me at eric@seod.com with your restaurant name and your three most recent photos. I will tell you which one a stranger scrolling a map would actually tap, which of the required shots you are missing, and which existing photo I would delete. Three or four lines back, no deck.

I ran restaurants for sixteen years and I answer these myself. If your photo set is already doing its job, I will say that instead of selling you a shoot.

Or keep reading more on restaurant marketing.