Image Alt Text for SEO: How to Write It Right

TL;DR: Image alt text exists primarily for accessibility. Describe what the image communicates, keep the wording succinct, include a keyword only when it naturally describes the image, and use an empty alt="" attribute for decorative images. Its SEO value comes from helping Google understand an image, not from creating another place to repeat a target query.

An “alt tag for SEO” is slightly awkward terminology. Alt text is technically an attribute on an image tag, not a separate tag. People still call it an alt tag, Google understands the query, and correcting everyone is not a productive use of your Tuesday (I use the shorthand in conversations too).

Alt text has two jobs, but they are not equal

The primary job is accessibility. WebAIM explains that “Screen readers announce alternative text in place of images, helping users with visual or certain cognitive disabilities perceive the content and function of the images.” If someone cannot see your image, its alt text supplies information or function they would otherwise miss.

This is not merely an SEO recommendation. WCAG 1.1.1 Non-text Content requires non-text content presented to users to have a text alternative serving the equivalent purpose. It is a Level A success criterion, the baseline conformance tier.

The second job is helping search engines understand the image. Google Search Central puts the relationship clearly:

“Google uses alt text along with computer vision algorithms and the contents of the page to understand the subject matter of the image.”

That sentence is the useful limit on SEO claims. Google does not say that inserting a target keyword into every image creates a predictable rankings increase. It says alt text is one source of understanding, alongside computer vision and the rest of the page.

Accurate alt text can support image discovery and give Google better context. I would not sell it as a standalone page-ranking lever. The distinction matters because writing “for Google” is how a useful description turns into a list of synonyms nobody would want read aloud.

Write first for the person who cannot see the image. If the resulting description naturally contains a relevant search phrase, fine. Usually the accessibility and SEO objectives align. Stuffing begins when they do not, and someone tries to force the alignment anyway.

The practical rule: accurate, equivalent, succinct

WebAIM says alt text should “be accurate and equivalent in representing content and function” and “be succinct.” That is a better writing standard than counting characters.

The same image, done three ways:

<!-- Good: describes the image -->
<img src="/puppy.jpg" alt="Dalmatian puppy playing fetch in a park">

<!-- Bad: filename junk and keyword stuffing -->
<img src="/IMG_1234.jpg" alt="puppy dog pet cute puppy for sale buy puppy">

<!-- Decorative: intentionally empty so screen readers skip it -->
<img src="/divider.svg" alt="">

Ask one question: What would a person miss if this image disappeared?

If the answer is a product’s appearance, describe the relevant product details. If it is the action performed by an icon, describe the action. If it is a trend shown by a chart, identify the chart and put the full finding or dataset in nearby copy. If the answer is nothing, the image is probably decorative.

Google Search Central’s image guidance gives a useful progression:

The final version works because every word contributes. It identifies the breed, subject, and action without becoming a miniature essay.

Describe this specific image

“Blue running shoes beside a gym bag” is more useful than “shoes.” “SEO dashboard showing declining organic clicks” is more useful than “SEO graph.” Specificity helps a screen-reader user understand what this particular image contributes, while giving the crawler less ambiguous information.

Do not reuse one description across every product angle or screenshot. If three photos show the front, sole, and side of a shoe, each should describe its own view. Duplication may be convenient, but it is not equivalent to the individual image.

I have failed at this with automated content: one acceptable template gets copied across a batch, then nobody checks whether it still describes image 47. Automation removes typing. It does not remove the need for correct decisions (an inconvenient distinction, but a real one).

Include a keyword only when it belongs

If a photo genuinely shows a brown leather laptop bag, brown leather laptop bag with front pocket is accurate alt text. The keyword overlap is natural. If the image shows a courier carrying an unmarked cardboard box, adding “leather laptop bag” because the page targets that phrase is not optimization. It is a false description.

If the terminology is still fuzzy, our guide to what a keyword is separates a target query from language that belongs naturally in content. Apply the same test here: would you use these words if search volume were hidden from you?

Google’s warning is stronger than the shortened version often quoted:

“Avoid filling alt attributes with keywords (also known as keyword stuffing) as it results in a negative user experience and may cause your site to be seen as spam.”

You do not need five synonyms for one object. One accurate phrase is enough.

Skip “image of” and “photo of”

WebAIM recommends avoiding phrases such as “image of” and “graphic of” because screen readers already announce the content as a graphic. Write “Golden retriever sleeping under a desk,” not “Image of a golden retriever sleeping under a desk.”

There are exceptions where the medium is meaningful. A page comparing a photograph with an oil painting may need to identify which is which. The default is simpler: describe the content, not the existence of the file.

There is no official 125-character limit

You will see 125 characters repeated as if Google or WCAG carved it into a standard. Neither defines that hard cap in its published guidance. It is a rule of thumb associated with older screen-reader behavior, not an official requirement.

Use WebAIM’s actual standard: succinct without sacrificing accuracy. Most ordinary images need a short phrase. A complex chart may require much more information, but that information usually belongs in the page copy rather than inside an enormous alt attribute.

The W3C WAI alt decision tree recommends including the information contained in a complex image elsewhere on the page. That also benefits sighted readers who need the figures in a searchable, selectable form.

Decorative images need empty alt text, not descriptive filler

If an image adds no meaning, use an empty alt attribute: alt="".

Examples include visual dividers, background flourishes, and an icon placed beside text that already communicates the same thing. WebAIM states: “Decorative images should have alt="".” WCAG likewise says pure decoration must be implemented so assistive technology can ignore it.

An empty attribute is not the same as a missing attribute:

• alt="" means the image is intentionally decorative and should be skipped.
• No alt attribute leaves assistive technology without that instruction. Depending on the screen reader and configuration, the filename may be announced instead.

That can turn a harmless divider into “blue-line-final-v3-dot-png” read aloud between paragraphs. An empty attribute is the fix, not another description (one of the few optimization tasks where less output is better).

The difficult part is classification. A shopping-cart icon beside visible “Add to cart” text is redundant and may be decorative. The same icon used alone as a clickable control is functional, so its text alternative should communicate the action.

This is why a scanner reporting 300 empty alt attributes has given you a queue, not a diagnosis. From what we see across sites using SEOJuice, a meaningful portion of empty or missing-alt cases require a role decision before text should be generated. Some images need descriptions; others should remain silent. Blindly filling every field can make accessibility worse while producing something that looks “100% optimized” in a report.

Context can change the right answer

The same image file can need different alt text on different pages. WebAIM makes the point under a section literally titled “Context is Everything.”

Consider a company logo:

• In a linked site header, its purpose may be navigation, so “Acme home” communicates the destination.
• In a case study comparing visual identities, “Acme logo with blue mountain symbol” may describe relevant visual information.
• Beside the company name in text, it may add nothing and can be decorative.

The file itself does not know its role. A filename-to-alt-text script does not know either, unless it can interpret the page context and element function.

Product-heavy sites expose this problem quickly. A listing may include front, side, interior, detail, and lifestyle photos. Five repetitions of “brown leather laptop bag” conceal the very distinctions the images were added to show. Marketplace discovery also involves more than alt text; our Etsy SEO guide covers that broader system.

Alt text is only one part of image SEO

Perfect alt text cannot rescue an image Google struggles to discover, render, or associate with relevant content. Google’s image guidance also covers filenames, placement, quality, sitemaps, supported formats, lazy loading, and structured data.

Use descriptive filenames and relevant placement

Google recommends filenames that are “short, but descriptive,” giving my-new-black-kitten.jpg as better than IMG00023.JPG. Treat this as useful information architecture, not evidence of a large or precisely weighted ranking factor.

Google also says, “Wherever possible, make sure images are placed near relevant text and on pages that are relevant to the image subject matter.” A product diagram beside its explanation supplies clearer context than the same diagram stranded in a general gallery.

Do not rename image URLs casually on an established site. Changing a filename changes its URL unless your media system abstracts that detail, so migrations may require redirects and updated references. A cleaner filename is not worth creating broken images or throwing away an indexed URL (I have seen tidying projects become migration projects surprisingly fast).

Make lazy-loaded images discoverable

Lazy loading is compatible with search, but implementation matters. Google says relevant content should load when it becomes visible in the viewport and warns against requiring clicks or scrolling because “Google Search does not interact with your page.”

Its practical test is straightforward: if an image URL appears in the src attribute of the rendered image element, the setup works correctly. Check the rendered HTML after JavaScript, theme, or plugin changes, particularly on image-heavy mobile pages. Our mobile SEO guide covers the performance side of that decision.

I no longer assume a lazy-loading library works because the page looks correct in my browser. During SEOJuice’s migration from seojuice.io to .com in January 2026, we were reminded that what a person sees and what a crawler can reliably retrieve are related, but not identical. Rendered output wins that argument.

Use supported formats and discovery mechanisms

Google Search supports BMP, GIF, JPEG, PNG, WebP, SVG, and AVIF images referenced through image elements. Format choice should balance quality, browser support, transparency or animation needs, and file size. The newest format is not automatically the right one.

Google also says that “High-quality photos appeal to users more than blurry, unclear images.” Compressing an image until product details disappear is not a performance victory. If images are difficult for Google to discover through ordinary crawling, an image sitemap can provide URLs it might not otherwise find.

Structured data is separate from alt text. Valid product or recipe markup can make images eligible for corresponding rich-result experiences, but an attribute added by a plugin is not proof that the complete markup is valid. Use our guide to testing rich snippets before relying on it.

A useful alt-text audit is a classification process

1. Find images with missing alt attributes. Keep missing attributes separate from intentionally empty ones.
2. Classify each image. Decide whether it is informative, functional, an image of text, complex, or decorative.
3. Identify its purpose on this page. The same file may perform a different job elsewhere.
4. Write the minimum equivalent description. Communicate the information or action a non-sighted user needs.
5. Remove stuffing and redundant prefixes. Delete keyword lists, “image of,” and wording that adds no meaning.
6. Review filenames and surrounding copy. Use descriptive filenames where practical and keep images near relevant text.
7. Inspect rendered lazy-loaded images. Confirm their URLs appear in rendered image elements.
8. Review charts and diagrams manually. Make sure the underlying conclusion or data is also available in page content.

Prioritize meaningful images on high-value pages rather than optimizing by raw count. A missing description on a product’s primary image matters more than an empty attribute on a decorative footer flourish. Audit tools tend to flatten both into one issue total; your remediation plan should not.

Where automation earns its keep, and where it stops

Writing one description is easy. Finding thousands of existing images, applying appropriate descriptions, and catching new uploads is repetitive operational work.

SEOJuice is a two-person team, Lida and me. We cannot hand-audit every image element across every customer site, and neither can most small site teams. That constraint is exactly why we automate the repetitive pass rather than pretending another downloadable spreadsheet solves execution.

SEOJuice can continuously write and apply descriptive alt text to missing images on a live site after installation through a JavaScript snippet or WordPress/CMS plugin. It is an execution tool, not a magic ranking button. Context still matters, decorative images still need to remain empty, and complex charts still deserve human review.

If missing alt text keeps reappearing in your reports, run the free SEO audit to locate affected pages. SEOJuice also has a free plan with no credit card, so you can test the automated pass on a real site before deciding whether it saves enough manual work.

Frequently Asked Questions

What is alt text, and is it the same as an “alt tag”?

Alt text is the alt attribute on an image element. Screen readers announce it, browsers may display it when an image fails to load, and search engines use it as one source for understanding the image. “Alt tag” is common shorthand, but technically it is an attribute rather than a tag.

Does alt text help SEO?

Yes, primarily by helping Google understand an image’s subject and context. This can support Google Images visibility and reinforce the relationship between an image and its page. Google does not describe alt text as a standalone page-ranking lever, and its primary purpose remains accessibility.

How do I write good image alt text?

Describe the content or function a person would miss if they could not see the image. Be specific and succinct, include a keyword only if it accurately describes the picture, and omit redundant prefixes such as “image of.” Functional images should describe their action or destination rather than their appearance.

What alt text should I use for decorative images?

Use alt="". This tells screen readers to skip the image. Do not omit the attribute entirely, because assistive technologies may handle a missing attribute differently and some may announce the filename.

Can alt text hurt SEO?

Keyword-stuffed alt text can harm user experience and may cause a site to be seen as spam, according to Google. Inaccurate descriptions also create misleading context. Over-describing decorative images is chiefly an accessibility failure: it adds noise without giving the user useful information.

How long should alt text be?

Keep it succinct without sacrificing accuracy. Google and WCAG do not publish an official 125-character cap. Ordinary images usually need a short phrase; detailed chart or graph information should be provided elsewhere on the page.

What else matters for image SEO?

Use short descriptive filenames, place images near relevant copy, choose a supported and appropriate file format, preserve useful visual quality, make lazy-loaded URLs available in rendered image elements, and consider image sitemaps for hard-to-discover files. Structured data may also make images eligible for certain rich-result presentations.