Photography for a small business website

Photography is the visual budget line owners cut first and regret longest. It is also the one part of a website that cannot be borrowed, which is exactly why it works.

A customer deciding whether to eat at your restaurant or let you into their house is looking for evidence that you are real and that the result will be good. A photograph of your dining room at seven in the evening is that evidence. A photograph of a generic smiling family is not.

What should I actually photograph?

Build a shot list before anyone picks up a camera. Five categories cover most businesses.

The place. Your exterior, so people recognize it from the street. Your interior, in the state customers see it. If you are a service business with a shop, that counts too.

The people. Faces of the people who will show up. This matters more than owners think, particularly for anyone entering a home. A team page with real faces converts differently than a team page with silhouettes.

The work. Whatever you make or do, shown in progress and finished. Trades should shoot completed jobs constantly. Restaurants should shoot the actual plates, not styled versions of them.

The details. The things you are proud of and the things that answer questions. The parking lot. The private dining room. The clean truck. The equipment.

The proof. Before and after, where that applies. It is the most persuasive image type available to a trade or a practice and it costs nothing but discipline.

Shoot for the pages you have, not for a gallery. Every page on your site should have at least one image that belongs to it specifically.

Does phone photography work, or do I need a professional?

A modern phone in good light produces perfectly usable web images. The variable that matters is light, not the camera.

Shoot near windows or outside in the first or last hours of daylight. Turn off overhead fluorescents, which make everything look institutional. Wipe the lens, which nobody does. Hold still and take five of each thing.

Where a professional earns their fee is people and interiors. Portraits that do not look like ID photos, and rooms that look like the room instead of a cave with a bright window. If your budget covers one shoot, spend it there and take the rest yourself.

What you should not do is mix. A homepage with two beautiful commissioned photographs and eleven dark phone snapshots looks worse than eleven consistent phone snapshots. Consistency reads as competence.

And whatever you shoot, keep the raw files somewhere you control, at full size. They are a business asset. If a photographer holds the originals and licenses them to you narrowly, you will pay again the next time you print a menu.

How much do unprocessed photos actually cost me?

Where does the speed argument get quoted wrong?

Constantly, and this is worth knowing because you will be shown these numbers by somebody selling you something.

There are at least three incompatible versions of Google's mobile speed statistic in circulation, and all three are Google's own. The February 2017 edition says that as page load time goes from one second to seven seconds, the probability of a mobile visitor bouncing increases 113%, and that 70% of pages were over 1MB. The February 2018 update of the same article says one second to ten seconds and 123%, and revises the page weight figures to the 79%, 53% and 23% quoted above. A third version, "one to three seconds, 32%," appears in neither primary document.

So when you see "70% of mobile landing pages" quoted as a current figure, you are looking at a 2017 number that Google itself replaced with 79% in 2018, usually detached from what it was actually measuring. The two bounce figures are not two findings either. They are one study, re-run, and presenting 113% and 123% side by side as independent evidence is a tell that whoever wrote the deck never opened the source.

The same care applies to the conversion side. Google, Deloitte Digital and Fifty-Five studied 37 brands across roughly 30 million sessions over four weeks in late 2019 and found that a 0.1 second improvement in mobile load time went with retail conversions 8.4% higher and, for luxury sites, page views per session 8% higher. Quote the baseline every time: one tenth of a second, against each site's own starting point, on 2019 data. A multiplier without a baseline is not a finding.

What do photos not do?

They do not affect your rankings. This is worth saying plainly because photography sometimes gets sold inside an SEO package, usually with a specific add-on attached.

That add-on is geo-tagging. Whitespark's Local Search Ranking Factors report, published 6 November 2025, had 47 local search experts score 187 factors, and presence of geo-tagged photos scored 23, the lowest score on the entire list. For scale, primary Google Business Profile category scored 227. Keywords in image metadata fields scored 30, and keywords in image filenames and alt attributes sit in the same low band. Grade it as expert opinion rather than a controlled test, but there is no evidence anywhere on the other side, and geo-tagging remains one of the most commonly sold local SEO extras.

The same 2026 report added something newer that is worth knowing before you get creative: presence of AI-generated photos or videos on the profile is now listed among the negative and suspension-risk factors, alongside keyword stuffing the description and AI-generated text. Generated images are not a shortcut to a photo library. They are a risk applied to the asset that produces most of your calls.

Good photos will not move you in the map results. What moves you there is a different set of factors entirely, and if you want to know where you actually stand, one ranking number will not tell you, which is why grid rank tracking exists instead of a single position.

Photos also do not fix an unclear offer. If a visitor cannot tell what you sell, better pictures of it will not help. That is a positioning problem, and it belongs to the brand decisions that come before a website.

What photos do is reduce doubt. They answer "is this real," "what will it look like," and "who am I dealing with." That covers three of the four things a hesitant visitor is worried about. The fourth is what it costs, and whether to publish pricing on your website is the decision that determines whether they ask or leave.

How do I keep photos from becoming a problem?

Two ways they go wrong, and both are common.

File size, which is the arithmetic above. Every image should be resized and compressed before it goes on the site. This is a five minute step that gets skipped constantly, and it is the single most common reason a redesigned site is slower than the one it replaced.

Age. Photos date faster than copy. Staff leave, the dining room gets refreshed, the truck gets replaced, the storefront sign changes. A site showing a team that has fully turned over is quietly telling customers you have not paid attention in years.

Put a recurring reminder on the calendar, twice a year, to look at every image on the site and ask whether it is still true. Add new job photos as they happen rather than in an annual push.

This matters most when the business itself changes. If you change your name, your look, or your location, the images have to move with it, and rebranding without losing the traffic you already have includes replacing every photo that shows the old sign.

Where does the case for photography break down?

Three places.

When nobody is arriving. Photography multiplies attention. It does not create it. A full library on a site with forty visits a month is a satisfying way to avoid the harder problem.

When the subject is genuinely invisible. Some professional services have no visual subject beyond the people. Forcing it produces the staged look that made everyone stop trusting stock photography in the first place.

When better photos make the page slower than the gain. A gallery of thirty images on a service page can push a passing Largest Contentful Paint into a failing one. The fix is fewer, processed, and loaded properly, not fewer megapixels in the camera.

What to do this week

Write the shot list. Walk your own space with your phone and note every shot you need, by page. Ten to twenty is normal.

Pick a morning with good light and shoot the easy categories yourself: the place, the work, the details. Take more than you need.

Book a photographer for the people, if you can. A short session covering headshots and two or three interior shots is usually enough.

Then process what you have. Run the weight arithmetic above, resize, compress, name the files something descriptive, and write one honest sentence of alt text for each.

Then put them on the site and take a speed reading before and after. New photography is one of the changes worth watching closely afterward, and what happens in the 90 days after a launch is the window where you find out whether anything changed.

Questions about what to shoot?

Email me at eric@seod.com with your business type and a couple of the photos you are planning to use. I will tell you which ones to keep, which ones to reshoot, and what is missing from your shot list based on the pages you have. It is a five minute reply and it usually saves a wasted afternoon.

More on the visual side in the web design and branding library.