How to Optimize Images for Local SEO

Here is the part most image-SEO guides get wrong before they get anything right: the single tactic they push hardest, geotagging your photos, is the one that does the least. Meanwhile the levers that actually work are unglamorous. Real photos instead of stock. Files that load in a fraction of a second because WebP is 25 to 35 percent smaller than JPEG (Google, web.dev). Captions a human being would find useful. This page is the hands-on version: the exact order to work in, the specific settings to use, and an honest note on how much each step is worth.

If you are still fuzzy on how local ranking works at all, read what is local SEO and how it works first; it covers the relevance, distance, and prominence model this page assumes you already know. What follows is narrower and more practical. It is the photo-by-photo workflow a Brantford business owner can run on a Saturday morning without hiring anyone, plus a clear list of which widely-repeated image tactics are worth your time and which are not.

What Google actually reads from an image

Start here, because it kills half the bad advice on its own. Google's official image documentation is explicit about what it uses to understand a picture: the file name, the alt text, the text surrounding the image on the page, structured data, and the image sitemap that helps it get discovered (Google Search Central). That is the list. It says nothing about GPS coordinates buried in the file, nothing about a location stamp, nothing about metadata you cannot see. When you optimize an image for local search, you are working on those documented inputs and on how fast the file loads. Everything else is decoration or superstition.

Two of those inputs carry most of the weight. The first is relevance: does the picture, its name, its caption, and the words around it match what the page is about and where you work? The second is speed: does the image slow the page down enough to hurt the experience Google measures through Core Web Vitals? Image work that improves relevance or speed is real. Image work that touches neither, no matter how many blog posts recommend it, is time you could have spent collecting a review.

The image optimization workflow, in order

Run these six steps on every image before it goes live. The order matters: sourcing and naming come before compression, because there is no point compressing a stock photo you should not be using in the first place. Times are rough, per image, once you have done it a few times.

Notice what is not on the list: no geotagging, no metadata stuffing, no keyword-loaded file names with five cities in them. The next three sections go deeper on the steps that carry the most weight, then a section explains, with sources, why the missing tactic is missing.

Real photos beat stock, every time

This is the highest-value step and the one owners skip most. A photograph of your actual crew on an actual job in a recognizable part of town does three jobs at once: it gives a human visitor a reason to trust you, it gives Google original content that exists nowhere else on the web, and it demonstrates the first-hand experience that both quality raters and AI engines now weight heavily. A stock photo does the opposite of all three. It is on a thousand other sites, it proves nothing, and to a trained eye it quietly signals that nobody at the business owns a camera.

There is a local nuance here that generic guides miss. For a service business, the geographic relevance you want does not come from a hidden coordinate; it comes from the visible content of the picture and the words around it. A roofing photo captioned with the street or subdivision, sitting next to a paragraph about that job, tells Google far more about where you work than any EXIF field ever could. That is the honest way to earn a local signal, and it is the version that survives Google stripping metadata on upload.

Naming and alt text: describe, do not stuff

File names and alt text are the two inputs you control by typing, and both reward plain description over keyword games. Google's guidance on alt text is direct: create useful, information-rich text that describes the image in the context of the page, and avoid stuffing keywords, which it treats as a spam signal (Google Search Central). The best alt text is the sentence you would say to someone who cannot see the screen. "Red brick bungalow with a new charcoal asphalt roof in Brantford" is useful to a screen-reader user, to Google, and to anyone whose image failed to load. "Roofing Brantford roof repair Brantford roofers near me" is useful to no one and actively risks a penalty.

File names follow the same logic. Rename the file before you upload it, use lowercase words separated by hyphens, and describe the actual subject. Include the city when it is genuinely true of the photo, and leave it out when it is not. One honest location reference in a file name and one in the alt text is plenty; repeating it in every field is the stuffing pattern the guidance warns against. If you want a formal way to add location context, structured data is the legitimate channel: the schema.org ImageObject type has a caption and a contentLocation property built for exactly this, unlike EXIF geotags, which Google does not read for ranking.

Size and format: the speed half of image SEO

Half of image SEO is not really about images at all; it is about page speed, which is technical SEO. Oversized images are the most common cause of a slow local business page, and slow pages lose both rankings and the impatient mobile searcher standing in a parking lot deciding who to call. The fix has two parts: serve the image at the size it actually displays, and serve it in a modern format. On the format question, Google's own web.dev guidance is clear that WebP images run 25 to 35 percent smaller than JPEG and PNG at equivalent quality, with Facebook reporting up to 80 percent savings on PNGs after switching. AVIF goes smaller still.

The practical recipe is short. Export each image close to its display dimensions rather than uploading the full-resolution original. Convert to WebP or AVIF and provide a JPEG fallback with the <picture> element so older browsers still get an image (Google Search Central). Aim to keep most in-body images under roughly 150 kilobytes, and reserve the larger budget for a single hero image. Add native lazy loading so images below the fold do not block the first paint. None of this requires a developer; most website builders and a free tool like Squoosh handle it. Where image speed fits alongside your other technical priorities is something we lay out in content versus technical SEO for small business.

The EXIF geotag myth, corrected

Now the tactic this page exists to correct. A large share of local-SEO advice tells you to geotag your images: embed GPS coordinates into each photo's EXIF metadata so that Google supposedly learns where your business is. It is presented as a secret edge, sometimes as "the ultimate proof of local relevance." It is not an edge, and here is the evidence, not just an opinion.

First, Google's own image documentation lists everything it reads from an image, and EXIF geotags are not on that list (Google Search Central). Second, Google Search Advocate John Mueller has stated plainly that there is no need to geotag images for SEO (Search Engine Land, 2025). Third, images routinely have their EXIF data stripped during processing, so the coordinate you carefully embedded often does not survive to be read at all. Fourth, and most telling, the one study that local blogs cite to defend the tactic actually undercuts it: across 27 businesses tracked over ten weeks, one improved, six saw no change, and four of those six dropped in the rankings after geotagging (Search Engine Land, 2025). That is not a lever. That is noise with a slight downward tilt.

To be fair to the nuance, a couple of small experiments have reported a narrow effect for hyper-local "near me" queries, and that ambiguity is exactly why the myth persists. But weigh it honestly. The upside is unconfirmed, contested, and tiny; the metadata is frequently discarded; and Google's public position is that it is unnecessary. Set that against the fifteen minutes it takes to geotag a gallery, and the tactic loses on effort alone. Every one of those minutes returns more if spent compressing a hero image or writing a real caption. This is the difference between Engineering Principles and folklore: we recommend the inputs Google documents and can demonstrate, and we tell you when a popular tactic is folklore, even when saying so is less exciting than selling a secret.

Google Business Profile photos are a separate surface

Everything above concerns images on your website. Your Google Business Profile photos are a different job with different rules, and for a local business they are often the more visible of the two, because they appear directly in the map pack and the profile panel where people choose who to call. Google has not confirmed that profile photos are a direct ranking factor, so treat them as a conversion and trust surface first. A searcher comparing three plumbers will click the one whose photos look like a real, active business over the one with an empty gallery or an obvious stock header.

Google's own specifications for profile photos are concrete, so match them. Upload JPG or PNG files between 10 KB and 5 MB, at a recommended 720 by 720 pixels, with 250 by 250 the practical minimum (Google Business Profile Help). Keep them in focus, well lit, and free of heavy filters or added text, which Google explicitly discourages. Cover the useful categories: a clean logo, a strong cover image, real interior and exterior shots so people recognize the place from the street, the team, and your finished work. Then keep a steady trickle going; a few genuine new photos each month reads as a live business far better than a large gallery uploaded once and never touched. New uploads can take 24 to 48 hours to appear.

If you want a structured way to work through the profile side, our Google Business Profile optimization checklist covers photos alongside the categories, services, and review habits that carry more weight than any single image ever will.

How to tell it is working, and how much to expect

Set expectations honestly, because image work is a small lever pulled well, not a switch. On its own, optimizing your photos will rarely lift you several spots in the map pack; how far it moves you depends on how competitive your city and category are and on how strong your reviews and proximity already are. What it reliably does is remove drag and add trust: faster pages, clearer relevance, and galleries that convince a human to choose you. Those effects compound with the rest of local SEO rather than replacing it.

Measure the things image work actually touches. Watch your Core Web Vitals and mobile page-speed scores in Search Console before and after a compression pass; that delta is immediate and real. Watch image impressions and clicks in the Search Console performance report filtered to the image tab. On the profile side, watch the photo views and the calls and direction requests Google reports, and note whether your listing starts winning clicks against comparable competitors. If you would rather have the whole picture mapped for you, that is the first vector of our local SEO service, and the broader reasoning behind how we prioritize this work lives in our research library.