Photo metadata and SEO: what actually counts

There is a lot of confident advice about photo metadata and search rankings, and most of it is folklore. This is what Google's own documentation says, field by field: what helps you rank, what only changes how your images are displayed, and what does nothing for search at all but is still worth doing. Every claim here is checked against the source and linked.

The short version, before the detail

Most articles about photo metadata and SEO treat every field as equally important. They are not. Here is the order, from most useful to least, based on what Google publishes rather than on what gets repeated in Facebook groups.

1. The words on the page around the image. Captions, headings, and the paragraph the photo sits inside.
2. Alt text. Google calls this the most important attribute for describing an image, and writing it is also the accessible thing to do.
3. Structured data and an image sitemap. These change how your images get found and displayed rather than how highly they rank.
4. The file name. Helpful, free, and smaller than you have been told.
5. Embedded IPTC fields. Real value for credit and licensing. No documented ranking effect.
6. EXIF camera data. Useful to you. Not a search signal.

If you want the wider picture, keywords, pages, local search and the rest, that lives in our photography SEO guide. This article covers only the image files themselves, which is the part photographers most often get wrong.

What Google actually documents

Google publishes an image SEO best practices page, and it is short enough to read in ten minutes. It is worth doing, because the gap between what it says and what the internet says it says is wide.

On file names, Google's wording is careful: the filename "can give Google very light clues about the subject matter of the image." Very light. On alt text it is much stronger, calling it "the most important attribute when it comes to providing more metadata for an image," and explaining that Google uses it "along with computer vision algorithms and the contents of the page" to work out what the picture shows.

That last phrase is the one to hold on to. The image is read in the context of the page. Google states plainly that it "extracts information about the subject matter of the image from the content of the page, including captions and image titles," and it tells you to place images near relevant text on relevant pages.

Summarized from Google's image SEO documentation and its image license metadata documentation, checked August 2026.

File names: real, but small

Descriptive file names are worth doing, for two reasons that have almost nothing to do with each other. The search reason is the light clue above. The bigger reason is that you have to live inside this library for the next decade.

A name like IMG_4471.jpg tells nobody anything. elena-marcus-brooklyn-winery-first-look.jpg tells a search engine a little and tells you a lot. Use lowercase, put hyphens between words rather than underscores or spaces, and keep it readable rather than stuffed. A file name that reads like a list of keywords looks exactly like what it is.

This only matters for the small number of images you actually publish: portfolio pages, blog posts, your homepage. Renaming your whole archive to please a search engine is wasted effort. Naming your archive well so you can find a shoot in 2031 is not, and we set out a system for that in photo file naming for photographers.

One caution before you spend a Sunday on this. Many website and gallery platforms rewrite file names on upload, which means your careful naming never reaches the web at all. Upload one image, then look at the URL it ends up with.

Alt text does the most work

Alt text has the highest leverage on this list and it takes about fifteen seconds per image. It is also, first of all, an accessibility feature: it is what a screen reader says out loud to someone who cannot see the picture. Write it for that person and the search benefit arrives on its own.

What good alt text looks like

• Weak: "wedding photo"
• Stuffed: "brooklyn wedding photographer best nyc wedding photographer wedding photos"
• Good: "Bride and groom laughing during their first dance at Brooklyn Winery"

The good one describes the picture in a normal sentence and happens to contain words a person might search for. Google's own instruction is to write "useful, information-rich content that uses keywords appropriately and is in context," and to avoid keyword stuffing. A run of city names is stuffing, and it reads as stuffing to a human too.

Two practical rules on top of that. Decorative images, a texture, a divider, a background pattern, should carry empty alt text so a screen reader skips past them instead of announcing noise. And do not begin every alt attribute with "photo of", because the screen reader has already said that it is an image.

EXIF and IPTC: credit, not ranking

This is where the folklore lives, so it is worth being precise. Embedded metadata is the information written inside the file itself: EXIF for camera settings, IPTC for the human fields like creator, copyright, caption and keywords. The full tour of what each field is for is in our photo metadata guide.

Here is what is true. Google does read certain IPTC fields, and it has a documented feature built on them. Its image license metadata documentation lists Copyright Notice, Creator, Credit Line, Licensor URL and Web Statement of Rights, and explains that these can surface as image credits and make a picture eligible for a Licensable badge. Google states the advantage of embedded data plainly: it "is embedded into the image itself, and the image and metadata can move from page to page while still staying intact."

Here is what is not true. Nowhere does Google say embedded metadata affects ranking. It is an appearance and attribution feature, and Google says outright that it does not guarantee the information will be shown at all. Filling your IPTC keyword field with search terms will not move you up a single position.

Fill those fields anyway, for the right reasons: credit that travels with the file when somebody reposts it, a copyright line that makes a takedown request easier, and a keyword system that makes your own archive searchable. That last one is worth far more to you than it is to Google, and we go through it in keywording your photos.

One more thing to check. IPTC itself warns that content management systems, digital asset managers and content delivery networks routinely strip embedded metadata on upload, and its FAQ on Google Images names tools like WordPress, Cloudflare and Cloudinary among the usual culprits. Download one image back from your live site and inspect it. If your fields are gone, that is a platform setting, not a metadata problem.

Five beliefs that do not survive the documentation

Each of these gets repeated with total confidence, and none of them holds up against the source.

"Geotagging your photos helps you rank locally"

Google's image guidance says nothing about GPS coordinates in EXIF. Local ranking runs on your pages, your business profile and your reviews, not on the coordinates written inside a JPEG. Put the location in the caption and the alt text where it can actually be read, and put the real effort into your Google Business Profile.

"Stuffing IPTC keyword fields gives you an edge"

There is no documented ranking use for it, and no human ever sees it. It is effort spent in a room where the meter is not running.

"You must strip all metadata for page speed"

IPTC and EXIF fields are a few kilobytes against a photograph of several hundred. If your pages are slow, it is because you are serving a 4,000 pixel wide image into an 800 pixel wide slot. Fix the dimensions and the format first.

"Alt text is where the keywords go"

Alt text is where the description goes. The keywords turn up by themselves when the description is accurate.

"More images on a page means more chances to rank"

More images means a heavier page and more descriptions written badly at midnight. Fewer, better chosen images with real text around each one does more work. Choosing them is its own skill, and it is the subject of selecting for a cohesive portfolio.

The page around the image beats the file every time

If you take one thing from Google's documentation, take this: the file is read in the context of the page. A perfectly named, perfectly tagged image sitting alone in a gallery grid with no words near it gives a search engine almost nothing to work with. The same image inside a post about that venue, with a caption underneath and a paragraph around it, gives it plenty.

This is the real argument for writing about your shoots rather than only posting them. A gallery grid is a wall of files. A post about the wedding hands every image a caption, a venue name, a season and a sentence of context, which is exactly the material Google says it uses to understand a picture.

Three technical items are worth an afternoon on top of that:

• Serve images at the size they are shown. Google notes that images are often the largest contributor to overall page size and points to responsive techniques, srcset and the picture element, with an ordinary src as a fallback.
• Submit an image sitemap if your images sit on a CDN or load through scripts. Google's stated purpose for it is finding images it might not otherwise discover, which is a genuine problem on gallery platforms.
• Add structured data where it fits your pages. It does not lift your position, it changes what your result is allowed to look like, including a badge in Google Images.

Most photography sites lose far more traffic to slow pages and thin text than to any metadata field, and we listed the usual offenders in photography website mistakes.

Where this fits in a normal week

None of this should turn into a second job. The trick is doing each piece at the moment you are already handling the file, and only for the images that will go online.

At ingest, set your copyright and creator fields once as a preset, so every file carries them from that moment onward. That is a two minute setup that runs forever, and it is the part of the metadata story with the clearest payoff, because credit travels inside the file even when somebody crops your watermark off.

Then, when a shoot is picked and you choose the eight or ten frames that will go on the site, do the publishing work on those and only those: a readable file name, an honest alt sentence, a caption with the venue and the year. Ten images at a minute each is ten minutes. Ten thousand images is a project you abandon in week two, which is how most photographers end up with no metadata anywhere.

The selecting comes first, and it is the slow part. Kepla is being built to make that first pass on your iPhone or iPad on the drive home, then carry on to the Mac when a shoot needs a closer look. It picks a shortlist and never touches your files: nothing is deleted, nothing is moved, nothing is renamed, so the naming and metadata scheme you just built stays exactly as you set it up. The picking app is still in development. The booking page is live and free today.

For the strategy that sits above all of this, go back to the photography SEO guide and start with the pages rather than the pixels.

FAQ