Method
How Suede Audit calculates a readiness score.
Each audit fetches one public page, the homepage you submit, plus three discovery files already public on that domain: robots.txt, llms.txt, and sitemap.xml. Suede Audit does not log in, does not crawl beyond those four documents, and does not run prompts inside ChatGPT, Perplexity, Gemini, or any other answer engine.
How the score is weighted
Each check carries a severity: high, medium, or low. High-severity checks count five times toward the score, medium-severity checks count three times, and low-severity checks count once. The readiness score is the percentage of weighted checks a site passes, rounded to the nearest whole number.
The 28 checks add up to 90 points. The check reference lists every one with its severity, its exact pass rule, and the repair for a failure, grouped into page access, crawler access, discovery files, structured data, metadata, content structure, and retrieval chunk size.
- 90 and above — A
- 80–89 — B
- 70–79 — C
- 55–69 — D
- Below 55 — F
Three pillars
Each check rolls up into one of three pillars, each scored the same way as the overall report:
- Discoverability — reachability, crawl policy, canonical routing, and the public discovery files (robots.txt, llms.txt, sitemap.xml).
- Entity evidence — structured identity data: JSON-LD presence, organization type, and verified identity links.
- Content clarity — page meaning, supporting content, share metadata, and response weight.
What the score does not prove
A high score is not proof of an AI citation. Buyer-prompt evidence requires separate, timestamped runs inside each answer engine, which this audit does not run. The score measures whether an answer engine can technically reach and parse a page — not whether it will recommend, rank, or cite it.
Image alternatives and page evidence
The image check tests whether each image declares an alt attribute. Empty alt text can be correct for decorative images; the report counts empty, non-empty, and missing attributes separately and flags empty alternatives for manual review. It does not certify that descriptions are accurate or that the page is accessible. See the W3C guidance on decorative images.
The evidence summary reports title, description, main heading, language, canonical declaration, schema types, and entries in the root sitemap. These are page declarations, not verified business facts. A sitemap index counts child sitemaps, not their pages; no sitemap count proves indexing.
Limits and caching
Each browser gets one free audit. The finished report is cached in that browser's local storage, so its report link can reopen without using another scan. Suede Audit accepts public HTTP and HTTPS URLs only; it will not inspect private, internal, or non-routable addresses.