Your answer lives inside an image — invisible to the box

Your answer lives inside an image — invisible to the box
Spotted in the wild — an infographic-heavy page with the actual steps baked into a JPG, zero extractable text, losing the list snippet to a plain-HTML competitor.

Reportedly the extractor can't read text in images for snippet selection, and alt text alone won't carry a multi-step answer.

The fix — every answer that matters lives in real HTML text. Keep the graphic as a supplement, mirror its content in a proper <ol> or paragraph underneath.

Pretty doesn't get parsed. Watching this.