Keywords for images: how the right tags get your photos found
Published March 9, 2026
Every image you upload to a stock platform, website, or online store competes with millions of others. The thing that determines whether someone finds yours is the keywords attached to it. Not image quality. Not composition. Keywords.
That sounds reductive, and it is. A terrible photo with perfect keywords will still be a terrible photo. But a great photo with poor keywords might as well not exist. Buyers can't purchase what they can't find.
What image keywords actually do
When someone searches "woman working laptop cafe" on Shutterstock, the platform doesn't look at every image to see which ones match. It looks at the keywords, titles, and descriptions that photographers attached to their images. If your photo of a woman working in a cafe doesn't include those keywords, it won't show up. Simple as that.
The same principle applies beyond stock photography. Google Images uses alt text and surrounding text to understand photos. E-commerce platforms use product tags to surface relevant items. Blog posts with well-described images rank better in image search. Keywords for images matter everywhere images appear online.
The two types of image keywords
Descriptive keywords
These describe what's literally in the image. Objects, people, colors, settings, actions. A photo of a golden retriever running on a beach gets: golden retriever, dog, beach, running, sand, ocean, waves, pet, animal, outdoors.
Most people start here, and that's correct. Descriptive keywords are the foundation. Get these wrong and nothing else matters.
Conceptual keywords
These describe what the image represents or how it might be used. That same dog-on-beach photo could also be tagged with: freedom, joy, summer, vacation, active lifestyle, pet owner, happiness, play.
Conceptual keywords are where most photographers leave money on the table. A marketing team looking for a "freedom" or "joy" image might land on your beach dog photo, but only if you tagged it with those concepts.
How many keywords should you use?
It depends on the platform. Adobe Stock allows up to 25. Shutterstock takes up to 50. Getty has its own limits.
The common advice is to use as many as the platform allows, but that advice is wrong. Using 50 keywords where only 30 are relevant means 20 of your tags are noise. Stock platforms track click-through rates against keyword matches. If buyers search "mountain" and land on your photo but it's actually a hill, that hurts your ranking over time.
Use as many relevant keywords as you have. If that's 25, stop at 25. If you can genuinely fill 50, go for it. We wrote a whole piece on how many keywords a stock photo should have if you want the platform-by-platform breakdown.
Picking keywords that buyers search for
This is the hard part. You need to think like a buyer, not a photographer.
Photographers tag images with what they see: "bokeh," "f/2.8," "35mm." Buyers search for what they need: "business meeting," "happy family," "healthy food." The overlap between these two perspectives is surprisingly small.
A practical exercise: look at your photo and ask "who would pay to use this image, and what would they type into the search bar?" A marketing manager at a health insurance company searching for website imagery types very different things than a photographer describing the same photo.
Common keyword mistakes
Being too generic
"Nature" and "beautiful" are almost useless as keywords. Millions of images have them. Your photo of a specific waterfall in Iceland competes with every sunset, flower, and mountain on the platform. Use "seljalandsfoss waterfall" or "iceland waterfall" instead. Specific keywords face less competition and attract more qualified buyers.
Keyword stuffing
Adding "business" to a photo of a coffee cup because someone might use it in a business context is a stretch. Platforms notice when buyers don't engage with your image after searching certain terms. Keep your keywords honest.
Ignoring the title and description
Keywords aren't the only metadata that matters. Your title and description provide context that helps platforms understand what the image is about. A good title is specific and natural: "Freelancer working on laptop in busy downtown cafe" tells the platform a lot more than "Person computer."
Using the same keywords for every photo
If you shot 20 photos at the same location, each photo still needs its own keywords. The wide establishing shot and the close-up detail shot serve different search queries. Copying the same keyword set across a batch is fast but it costs you discoverability.
Keywords for different types of images
Stock photography
Focus on buyer intent. What campaigns, articles, or projects would use this image? Tag the concept as much as the content. Include the platform's preferred keyword format (some want single words, some accept phrases). Our stock photo keywording guide covers this in detail.
E-commerce product images
Include the product name, category, material, color, use case, and any relevant attributes. "Leather crossbody bag brown women travel" is better than "bag" for a product listing.
Blog and website images
Use descriptive alt text that serves both accessibility and SEO. The alt text should describe the image for someone who can't see it. That description naturally contains the keywords that help it rank in Google Image search.
Using a keyword generator to speed things up
Writing keywords from scratch for every image is slow. A keyword generator analyzes your photo and produces a list of relevant tags, a title, and a description. You review, edit, and export.
The advantage isn't just speed. A generator often catches keywords you wouldn't think of, because it evaluates the image systematically rather than going with whatever words come to mind first.
KeywordPic generates up to 50 keywords per image and exports CSV files for Adobe Stock, Shutterstock, Getty, and 30+ other platforms. You can try it free with 10 generations.
If you're looking at different tools, we compared the options in our stock photo keywording tools comparison.