What an SVG file actually is
SVG is not a grid of pixels at all: it is an XML document describing shapes, paths and text, which the renderer draws at whatever size it is asked for. That is why a 4 KB logo stays perfectly sharp on a billboard, and why the file is often smaller than a thumbnail of itself.
SVGs come out of Illustrator, Figma, Inkscape and icon libraries. Logos, icons, charts and diagrams are the natural cases — anything built from shapes rather than captured from a camera.
What you get from WebP
WebP output uses lossy compression with a quality control, but unlike JPEG it can keep an alpha channel at the same time — one format for photographs and for graphics with transparency.
Every current browser reads WebP; Safari joined in version 14 in 2020, which was the last real holdout. Desktop applications are patchier, so it is a format to publish in rather than to archive in.
What changes when you convert SVG to WebP
This is where the file size actually drops. A SVG stores its image without discarding anything, and WebP's lossy mode decides what the eye will not miss — typically landing 25–35% below a JPEG of matching quality, and keeping the alpha channel while it does so, which JPEG cannot. The catch is not visual but practical: WebP is a format to publish in, and a poor choice for anything you are handing to someone else's desktop software.
Transparency survives intact. Both SVG and WebP carry a full 8-bit alpha channel, so cut-outs, drop shadows and anti-aliased edges convert without a white box appearing behind them, and the result still composites cleanly onto any background colour.
The important loss here is scalability rather than quality. An SVG is drawn fresh at whatever size it is asked for; the moment it becomes a WebP it is a fixed grid of pixels, and enlarging it later softens the edges. Rasterize at the largest size you will realistically need, and keep the original SVG — it remains the master copy.