Alt text for memes and text-heavy graphics

You just built a quote graphic in Canva — bold type over a solid color block, maybe twelve words that carry the entire point of the post. You get to the alt text field and your first instinct is to skip it. The words are right there on the image. Surely that's enough?

It isn't. And memes have the same problem, just funnier.

The text on the image doesn't count as text

A meme or quote graphic is a JPG or PNG. Whatever words you typed into the design tool got flattened into pixels the moment you exported it. A screen reader can't read pixels — it reads alt text, or it reads nothing. Without alt text, someone using a screen reader hears "image," or the filename, while the actual joke or quote sails by unread.

It's tempting to assume the platform handles this automatically. It mostly doesn't. Instagram and Facebook's automatic alt text is built on object recognition — it can guess "cartoon" or "text" or "one person, indoor," but it isn't a reliable transcription tool. It won't hand a screen reader user the actual caption on your meme, only a vague hint that some kind of image with some kind of text exists.

Write the words first, then the format if it matters

For a meme, that means transcribing the caption itself, not summarizing the vibe. "Funny meme about mornings" tells a screen reader user that a joke happened somewhere without them. Compare:

Notice the strong version does two things: it names the format briefly ("meme," "stick figure") so the visual style registers, then gives the exact caption in quotes. If a well-known meme format is doing a lot of the joke's work — distracted boyfriend, drakeposting, expanding brain — naming the format in a few words is often more useful than describing the picture itself, since regular viewers recognize the format instantly and a screen reader user deserves the same shortcut.

Quote graphics: the alt text basically is the text

Quote cards and text-heavy graphics are simpler in one way: there's usually no joke to set up, just a line you want read. The alt text should be close to the quote itself, not a description of the card it's sitting on.

If the graphic carries a long passage — a full paragraph, a stat-heavy pull-quote, a multi-line testimonial — you won't always fit every word before a screen reader starts truncating. When that happens, lead with the line that matters most and trim from there, the same way you'd tighten any other alt text that's running long.

Why this gets skipped anyway

Because at a glance, the image looks self-explanatory to whoever's publishing it — the text is right there, visibly, so it's easy to forget that "visible to me" and "readable by a screen reader" are two different things. The fix isn't complicated, it just requires remembering that every word baked into an image has to be typed again, by hand, into a completely separate field.

That's the same workflow tax that makes alt text get skipped everywhere else, just with an extra step: you have to actually reread your own meme text before you can describe it. Whatever tool you use to write alt text, treat any image with embedded words as a transcription job first and a description job second — the caption always comes before the color of the background.