Token bloat.

Token bloat is when a page's boilerplate (navigation, repetitive UI chrome, scripts, trackers, duplicated blocks) overwhelms the useful text that AI systems and search engines can actually use. SEODiff surfaces token bloat in your canonical /radar/domains/DOMAIN pages.

Docs: Token bloat metric · Example: /radar/domains/airbnb.com

What to look for

If token bloat is high, you usually have a “UI-heavy” DOM with relatively little main content. This is often worse on JS apps where the rendered DOM includes lots of repeated components.

Symptoms

Low word count vs very large DOM / HTML payload, repeated nav/footer blocks on every page, huge inline JSON, or content buried behind client-side rendering.

Fixes

Move primary content earlier in the DOM, reduce duplicated boilerplate, remove unnecessary inline data, and ensure server-rendered primary text exists without JS.

API

Token bloat is exposed in JSON reports under a dedicated field (name may evolve; use the report JSON as source of truth).

curl -s "https://seodiff.io/radar/domains/airbnb.com?format=json" | jq '.token_bloat_ratio'