Should I watermark my photos before or after resizing and compressing them?

The share is the same either way

Mark size in the tool is a share of the image width, so the overall look ends up similar:

So the question is not whether the mark will shrink, but what happens to it along the way.

Why resize first

• Cropping: the Image resizer has a mode that fills the new size and crops whatever overflows at the edges. Put the mark in a corner, then change the aspect ratio in that mode, and your mark may be cropped along with the edges. A mark added after resizing is placed inside the final frame, guaranteed.
• Sharpness: drawn on the final photo, the letters and their outline are rendered in the very pixels the customer will see. Shrunk along with the photo, the thin outline gets recalculated and blends into its surroundings, so small lettering looks less crisp.
• Speed: the tool draws every photo at full resolution. Thirty photos 1200 px wide are far lighter on a phone than thirty at 4000 px.

And why compress last

The watermark tool always saves JPEG and WebP at quality 92. Compress your photos to quality 70 and then watermark them, and the tool saves them again at 92: your compression is gone and the file may come out larger. Compressing as the final step with the Image compressor means the final quality is the one you chose.

A practical order

1. Resize each photo to the size your store needs. The resizer handles one image at a time.
2. Open the Image watermark editor tool, choose all the resized photos, up to 30, add the mark and download them as a ZIP.
3. Run the watermarked photos through the compressor at the quality you want, and check that the mark's letters are still clean.

One exception: if you need several sizes of the same photo, say one for the store and one for Instagram, watermarking the original once and then resizing is quicker, as long as you use a mode that does not crop the edges. Either way, keep the unmarked original.