Annotation exists to move a viewer's eye. Every mark you add either does that or adds noise, and most marked-up photos fail because they do too much: five arrows, three circles, and a paragraph of text competing for attention.
One photo, one point
If a photo needs more than three annotations, it usually needs to be two photos. Split a wide shot into a context frame and a detail frame; annotate the detail frame. Viewers understand a sequence far better than a cluttered single image.
Number your findings
When several items genuinely belong in one frame, use numbered badges on the image and put the descriptions in the accompanying text. Numbers stay legible when the image is compressed by a chat app; a small sentence on the photo does not.
Make annotations survive compression
- Use high-contrast styles: outlined text or a solid badge, never mid-grey on a mid-grey wall.
- Size marks as a fraction of image width, around 4–6%, so they hold up in a thumbnail.
- Avoid thin hairlines; they vanish first when a platform re-encodes the image.
- Export as JPEG at high quality, or PNG if the annotation has hard edges over flat color.
Say what, not just where
A circle tells the viewer where to look; it does not tell them what they are looking at. Pair every mark with two or three words — hairline crack, missing clip, wrong colorway. That combination is what makes an annotated photo usable without a phone call.
Keep the clean copy
Annotate a duplicate, always. The unmarked photo is the record; the marked one is the message. Anyone reviewing the file later will want both.
Do it now: open the free photo labeler, drop in your image, and drag the label where you want it. Everything runs in your browser — the photo is never uploaded.