What Is Alt Text, and How to Write Good Alt Text

Imagine reading a news article where every photo has been replaced by a blank box labeled IMG_2049.jpg. You would miss the story the pictures tell: the scale of a protest crowd, the expression on a player's face, the trend line in a chart. For millions of blind and low-vision people who browse with screen readers, that is the daily reality of the web. When an image has no useful alternative text, the picture might as well not exist.

Alt text, short for alternative text, is the fix. It is a brief written description attached to an image in a page's code. Screen readers announce it in place of the picture, browsers show it when an image fails to load, and search engines read it to understand what an image contains. It is one of the oldest accessibility techniques on the web, one of the simplest to get right, and still one of the most neglected.

When an image has no useful alternative text, the picture might as well not exist.

Why alt text matters

The scale of the problem is stark. WebAIM's 2026 Million report, which audits the home pages of the top one million websites, found that 53.1 percent of them had images with missing alt text. About 16.2 percent of all home page images lacked it entirely, roughly 10.8 images per page, and more than one in four images had alt text that was missing, questionable, or repetitive. The trend is improving slowly, down from 68 percent of pages in 2019, but the basics are still being missed on most of the web's most visited pages.

For screen reader users, the cost is direct. WebAIM's user survey found that NVDA and JAWS dominate on desktop while VoiceOver leads on mobile, and more than nine in ten respondents use a screen reader on a phone. These tools turn alt text into speech or braille. Without it, a screen reader announces the image's file name, or nothing at all. Worse, 45 percent of the images missing alt text in WebAIM's sample were linked images, which means the link itself becomes a mystery: the screen reader announces link, and the user has no idea where it goes.

Alt text also serves everyone else. It appears when images fail to load on slow connections, it feeds image search results, and it gives AI systems something accurate to work with when summarizing pages. Writing it is a thirty-second task with an outsized return.

What good alt text looks like

Good alt text answers one question: what would a sighted person take away from this image in context? Describe the content and the function, not every pixel. Context matters enormously: the same photo needs different alt text on a photographer's portfolio page than on a news article about a marathon.

Consider these before-and-after pairs. Bad: alt="IMG_2049.jpg". Good: alt="Golden retriever puppy sleeping on a blue rug". The file name tells nobody anything; the description conveys the image in eight words. Bad: alt="image of a chart showing data". Good: alt="Bar chart showing website traffic doubling from January to June 2026". Never start with image of or picture of; the screen reader already announces that it is an image. Bad: alt="click here" on a linked logo. Good: alt="Download the 2026 annual report (PDF)". For linked images, describe the destination or action, not the picture. Bad: alt="" on an informative photo (or no alt attribute at all). Good: an actual description, reserving empty alt text strictly for decoration.

A useful test: read your alt text aloud and ask whether a listener would understand why the image is there. If the surrounding paragraph already says the same thing, shorten the alt text or mark the image decorative. If the image is a complex chart, keep the alt text to a one-sentence summary and put the full data in the adjacent text where everyone can use it.

The most common mistakes

WebAIM found that 10.8 percent of images with alt text had descriptions that were questionable or repetitive, and the patterns repeat everywhere. The greatest hits: alt="image", alt="graphic", or alt="photo", which add nothing; pasting in the file name, which produces announcements like DSC underscore zero four seven two dot jpeg; repeating the caption word for word, which makes screen reader users hear everything twice; and stuffing keywords for SEO, which reads terribly aloud and can backfire in search rankings.

Two subtler mistakes deserve attention. The first is describing the image instead of its message: alt="A man in a suit standing at a podium with a flag behind him" is weaker than alt="The mayor announcing the new transit plan at city hall", because the second tells the reader why the photo matters. The second is forgetting that decorative images need alt="" with nothing between the quotes. A missing alt attribute is not the same as an empty one. Empty tells the screen reader to skip the image silently; missing leaves the software guessing, and it will often guess the file name.

How screen readers announce your alt text

Knowing the audience helps. On desktop, JAWS remains the most reported primary screen reader at 40.5 percent of WebAIM survey respondents, with NVDA close behind at 37.7 percent and VoiceOver on Mac at 9.7 percent. On mobile the picture flips: 91.3 percent of respondents use a screen reader on a phone, and VoiceOver on iPhone dominates at 70.6 percent. Most users run more than one screen reader, so alt text written to the standard works across all of them.

In practice, a screen reader pauses at the image, announces graphic or image, then reads your alt text in the page's reading voice. Braille display users feel it under their fingers instead. Keep descriptions under about 125 characters, since some screen readers truncate longer text, and write in the same language and tone as the surrounding page. Then test it yourself: every major screen reader is free to try except JAWS, and VoiceOver is already on your iPhone and Mac.

References

WebAIM Million 2026 report; WebAIM Screen Reader User Survey #10; Web Content Accessibility Guidelines (WCAG) alt text guidance; Apple VoiceOver documentation.