Alt text exists primarily so that people using screen readers can understand an image. That is reason enough to write it. It has a secondary effect worth knowing about: it gives the platform a plain-language description of what your image contains, which contributes to how your content is classified and, on search-oriented platforms, what it can surface for.
Two genuine benefits, one field, and almost nobody fills it in.
Why auto-generated alt text is not sufficient
Most platforms generate alt text automatically using image recognition. The output is generic by nature — something like "image may contain: person, indoor, table." It is technically accurate and practically useless: it identifies objects without conveying what the image is about.
For a screen reader user, that is the difference between knowing a photo exists and knowing what it shows. For classification, it is the difference between "food" and "cast iron skillet with seared steak, kitchen counter, home cooking."
How to write it
- Describe what is actually in the image, in a plain sentence. Not what you want it to mean — what a person would see.
- Lead with the subject. "A cast iron skillet with a seared steak" before the setting details.
- Include text that appears in the image. If your graphic has words on it, those words belong in the alt text — otherwise a screen reader user misses them entirely.
- Keep it to one or two sentences. Long alt text is tiring to listen to.
- Do not write "image of" or "photo of." Screen readers already announce that it is an image.
- Do not stuff keywords. A keyword list is unreadable aloud, which defeats the primary purpose, and platforms are not fooled by it.
| Poor | Better |
|---|---|
| Image of food | A cast iron skillet with a seared steak resting on a wooden board |
| #homecooking #steak #castiron #recipe | A close-up of a rare steak being sliced, showing the pink centre |
| My new setup! | A desk with two monitors, a mechanical keyboard, and a small plant beside a window |
| Chart | A bar chart showing hashtag reach dropping as post volume rises above 12 tags |
The graphic case, which matters most
If your post is a text-based graphic — a quote card, a statistics slide, a carousel with words on it — then without alt text a screen reader user gets nothing at all. The information exists only as pixels. This is the situation where filling in alt text moves from good practice to genuinely necessary.
It is also the case where the classification benefit is largest, because image recognition handles text-in-image poorly, and the platform otherwise has very little idea what the post says.
Where alt text goes on each platform
- Instagram. Advanced settings when composing, or edit an existing post and choose Edit alt text.
- Pinterest. The alt text field on the pin. Particularly worthwhile here given Pinterest’s search orientation.
- LinkedIn. An Alt button appears on images during composition.
- X. An ALT badge on attached images.
- Facebook. Edit the photo, then Alternative text.
- Video platforms. No alt text as such — the equivalent is accurate captions, which serve the same dual purpose of accessibility and indexable text.
A realistic habit
Writing alt text takes about fifteen seconds per image once it is a habit. The reasonable rule: always for graphics containing text, always on Pinterest, and as default practice everywhere else. It is the rare optimisation where the accessibility case and the discovery case point in exactly the same direction, which means there is no trade-off to weigh.