Resizing vs compressing: what is the difference, and when to use each
"Resize" and "compress" are often confused, but they change different properties of an image, and using the wrong one doubles your effort. This explains the difference and the order to use them when you need both.
They change two different things
Resizing changes the pixel width and height. A 4000×3000 image resized to 1000×750 has one sixteenth the pixel count and appears smaller on screen.
Compression changes the file's byte size while the pixel dimensions stay the same. A 1920×1080 image is still 1920×1080 after compression; it simply uses less data to store.
In short: dimensions decide how big the image looks, file size decides how heavy the file is. The two are related but not the same — a large-dimension image tends to be a large file, but you can compress the file size alone without changing dimensions.
When to use which
You only want a smaller file and the display size does not matter: compress. For example an email attachment that is too large, or an upload with a size limit.
You need specific pixel dimensions: resize. For example an avatar required at 200×200, a video cover at 1920×1080, or a web image kept within a certain width.
You need both — specific dimensions and a controlled size: resize first, then compress (see below).
The order to combine them
When a large image must be both a specific size and under a specific file size, the right order is resize first, then compress. Shrink the pixels first, and compression reaches a smaller file at the same quality, hitting a target KB more easily. Compressing first and shrinking afterwards means you encoded pixels you were about to discard — wasted work. This is also why the common reason for "it will not compress under 100KB" is that the dimensions were not reduced first.
More guides