Ready to Make Your Photo Blog Accessible?
Generate accessibility-friendly alt text for your whole archive in one session
Try PhotoScanr FreeFree to use • No sign-up required • Instant results
Writing Alt Text for Accessibility on a Photo Blog
By Duncan Rawlinson · Updated
Alt text is one of the smallest things you can change on a photo blog and one of the most consequential. For a sighted reader, an image either loads or it does not. For a screen reader user, alt text is the entire experience of your image. Skip it and you are effectively serving a blank space.
The good news is that writing useful alt text does not require formal accessibility training. It requires honest description, restraint, and a willingness to think about what a person who cannot see the image actually needs to know. This guide covers how to make those decisions for the kinds of images that show up on a working photo blog.
We will start with the WCAG 2.2 framework, then move into the practical question every photographer faces: which images need detailed descriptions, which need short ones, and which should have no alt at all.
What WCAG 2.2 Actually Requires
The Web Content Accessibility Guidelines, currently at version 2.2, are the standard most accessibility laws point to. The relevant rule for images is Success Criterion 1.1.1 Non-text Content. It requires that every non-text element either has a text alternative that serves an equivalent purpose, or is marked as decorative so that assistive technology can skip it.
Equivalent Purpose, Not Identical Description
The standard does not say "describe every pixel." It says the alternative should serve the same purpose as the image. A photo of a sunset on a personal blog post titled "The night I quit my job" probably needs different alt text than the same sunset on a meteorology site.
Ask yourself: what is this image doing on this page? Then write alt text that performs that function for someone who cannot see it.
Length Has No Hard Limit, But Norms Exist
WCAG does not specify a character limit. In practice, screen readers handle alt text best when it is between 20 and 150 characters. Beyond that, consider whether a longer description belongs in a figcaption or in the body text, with the alt acting as a quick orientation.
Conformance Levels and Why They Matter
WCAG defines three levels: A, AA, and AAA. Most laws and corporate policies require AA conformance. Image alternative text falls under Level A, which is the baseline. If you do nothing else for accessibility, getting alt text right is the minimum.
Decorative vs Informative Images
Not every image needs alt text. Some images are pure decoration and adding alt text actually makes the experience worse, because the screen reader interrupts the flow of reading to announce something irrelevant.
Decorative: Use Empty Alt
An image is decorative when removing it would not reduce the meaning of the page. Examples include background patterns, divider lines, hero images that already have a heading right above them, and icons that sit next to text labels saying the same thing.
For these, use alt="" with the empty quotes. This tells screen readers to skip the image entirely. Do not omit the attribute. An omitted alt attribute causes some screen readers to read the filename, which is rarely useful.
Informative: Write Real Alt
An image is informative when it carries meaning that is not duplicated in the surrounding text. Photos in a journal entry, product images, infographics, and any image that the post is "about" all qualify.
Write alt text that captures the core meaning. If the image shows a chart, the alt might describe the trend rather than every data point. If it shows a face, the alt might name the person and their expression rather than enumerate features.
What AI Gets Right and Wrong
AI vision models can generate plausible alt text in seconds, which is a huge productivity win. They are also confidently wrong in ways that matter for accessibility, and the patterns of failure are worth knowing.
What AI Gets Right
Modern vision models are excellent at identifying everyday subjects, common scenes, dominant colors, and basic compositional elements. For travel photos, food shots, product images, and most editorial content, they produce a draft that is accurate enough to publish with light editing.
What AI Gets Wrong
AI is unreliable for proper nouns. It will guess at the name of a building, a city, or a celebrity, and it will sometimes be very wrong. It tends to over-describe, including details that do not matter for the page's purpose. It struggles with cultural context, with images of disabled people that it sometimes describes in patronizing ways, and with specifying race or gender without making assumptions that may be wrong.
The Right Way to Use AI Alt
Treat AI output as a first draft. The model handles the tedious "describe the obvious" work in seconds, and you provide the one or two pieces of context the model cannot know: the location, the person's name, the reason this image matters in the post. That hybrid workflow is faster than writing from scratch and more reliable than trusting the AI alone.
Screen Reader Pacing
Screen readers read alt text inline with the rest of the page. The reading speed is usually faster than sighted reading. A 200-character alt that scans quickly on the page can feel like a digression on a screen reader.
Tips for Good Pacing
- Lead with the subject. Get to the noun fast so the listener knows what is being described.
- Avoid "image of" or "photo of." Screen readers already announce that an image is present. Repeating it wastes the listener's attention.
- Use natural punctuation. Commas and periods give the screen reader breathing points and make the description easier to parse.
- Be specific about what matters. "A grey tabby cat curled on a red velvet armchair" is more useful than "a cat sitting somewhere."
Alt Text vs Caption: They Are Different Jobs
The two get conflated all the time. Captions are visible to everyone and read by everyone. Alt text is read only by screen readers and shown only when the image fails to load. They have different purposes and should usually contain different content.
Captions Add Context Visible to All
A caption might say "Self-portrait, Sunday morning, 6 a.m. light." It assumes the reader can already see the image. It adds the human context that turns a photo into a story.
Alt Text Stands In for the Image
Alt text for the same image might say "A woman in a white robe stands at a kitchen window, sunlight cutting across her face." It tells the screen reader user what the image is.
When the Caption Already Describes the Image
If your caption is genuinely descriptive, the alt text can be empty so the screen reader does not say the same thing twice. This is a judgment call. When in doubt, give a brief alt that mentions the subject and let the caption handle the context.
How PhotoScanr Helps With Accessibility
PhotoScanr was built with the assumption that a working photographer will not stop to write 200 alt text strings by hand. The tool's job is to produce a strong first draft that a human can refine.
Style Preferences for Accessibility-First Output
You can configure a style preference that biases output toward concise, factual, screen-reader-friendly description. This avoids the AI's natural tendency to be flowery or to hedge with phrases like "appears to show" or "seems to depict."
Grounding to Reduce Hallucinations
The grounding feature constrains the model to describing what it can actually verify in the image. This dramatically cuts down on the AI inventing details that are not present, which is a frequent and harmful failure mode for accessibility text.
Batch Processing for Backlog Cleanup
If you have years of blog posts with empty alt attributes, PhotoScanr's batch mode can process them at scale. Pro covers 100 images per day in batches of 25, and Studio covers 600 per day in batches of 100. A weekend of focused work can clear most photo blogs.
Free tools for checking your work include the WAVE browser extension, axe DevTools, and the screen reader built into your operating system (VoiceOver on macOS and iOS, Narrator on Windows, TalkBack on Android). For SEO-side considerations alongside accessibility, see our photo SEO guide for Squarespace portfolios.
Putting It All Together
Accessible alt text is a craft of restraint. Decide whether the image is informative or decorative. If informative, describe what matters in the fewest words that capture the meaning. If decorative, leave the alt empty and let the screen reader move on.
AI tools can handle the volume problem and produce drafts that are usually 80 percent of the way there. Your job is the last 20 percent: the proper nouns, the human context, the editorial judgment about what this particular image is doing on this particular page.
Doing this well takes some practice. The first hundred alt texts you write will feel awkward. By the time you have written a thousand, the pattern becomes second nature. And the readers who depend on alt text will keep coming back to a blog that respects their time.