What a watermark actually does
It does two jobs. The first is attribution: anyone who sees the image knows whose it is, even when it has been reposted three times without credit. The second is deterrence: a photo with a mark across it is less attractive to someone looking for free material to pass off as their own. What it does not do is prevent theft outright — a determined person can crop or clone it out.
Corner or across the middle?
This is the central trade-off. A small mark in a corner keeps the photo clean and is the right choice for a portfolio, a shop listing or social media, where the image needs to look good. It is also the easiest to crop away. A repeated pattern across the whole frame is far harder to remove but visibly degrades the picture, which makes it suitable for proofs sent to a client before payment, not for public display.
Getting the opacity right
Around forty to sixty percent works for most images. Below thirty the mark disappears on busy backgrounds and stops doing its job. Above seventy it competes with the photo itself. White text sits better on dark or colourful images, black on light ones — check the corner where the mark will land, not the photo as a whole.
What to write
Shorter is better. A name, a domain or a copyright symbol with a year is enough. A domain has one advantage over a name: it tells the viewer where to find more, turning a stolen repost into a route back to you. Long sentences shrink to the point of being unreadable and take up space the photo needs.
Keep the original
Always watermark a copy. The mark is burned into the pixels and cannot be removed afterwards, so if you ever need the clean version — for a print, a client, a different crop — you will need the file you started with. Store originals separately and treat watermarked files as output only.