AI visibility (by Echorank)
What the AI-visibility category checks, which parts come from Echorank, what each plan includes, and what the results do and do not mean.
Updated
The AI visibility category is built with Echorank, ChatLogic Insights' AI-visibility product and EchoPilot's sister product. It answers two questions with evidence: can AI crawlers fetch and read your website, and, on Growth and Managed, does Claude name and describe your business correctly when asked about your service in your city. It does not predict or promise what any AI will say about you.
What is checked
| Rule | What it checks | Source |
|---|---|---|
| V1 | llms.txt is published | your site |
| V2 | Claude names the business in its answers to three questions built from your main service and city | Echorank (Claude) |
| V3 | Claude's description of the business matches your name, city and service | Echorank (Claude) |
| V4 | FAQPage structured data is present | your site |
| V5 | robots.txt lets the AI crawlers in; GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot and Google-Extended count first | Echorank |
| V6 | the homepage text is in the HTML (server-rendered), not only added by JavaScript | Echorank |
| V7 | every JSON-LD block on the homepage parses | your site |
| V8 | robots.txt lists the XML sitemap | your site |
| V9 | the homepage has a title, a meta description and Open Graph tags | your site |
| V10 | headings follow in order, without skipped levels | your site |
V7 to V10 only judge what Technical SEO does not, so nothing is counted twice: whether structured data exists is S7 and V4, whether a sitemap exists is S4, title and description quality is S1 and S2, the single H1 is S3. When there is nothing for V7, V8 or V10 to judge (no JSON-LD, no robots.txt, no headings), the rule is not assessable and the Technical SEO rule carries the finding.
The whole category weighs 4 of 100 in the overall score, as before; the ten rules share that weight.
What each plan includes
- Free score: a quick check of the homepage (no crawl): V5 and V6 from Echorank, plus V1, V4 and V7 to V10. The page shows the crawler list and Echorank's check bars. Echorank's content signals (About and FAQ pages) need a crawl, so the quick check leaves them out.
- Audit: the full check in every scheduled and manual audit, shown in the AI visibility section of the audit page and in the audit PDF. V2 and V3 are not assessable on this plan and are left out of the score.
- Growth: adds the answer checks V2 and V3, labelled "Claude (via Echorank)" with the date they were asked and an excerpt of each answer.
- Managed: adds a weekly AI-visibility benchmark of your homepage against the competitors you watch (up to five), shown on the competitors page and used, with its source, in the monthly market report.
What the answers mean
The answer checks ask Claude only. Echorank does not ask other AI assistants today, and EchoPilot does not say it does. An answer is one reading on one day: AI answers vary between askings and over time, so treat V2 and V3 as a signal to watch across audits, not a verdict. The fixes stay in EchoPilot's fix snippets and the 90-day memo.
When a check cannot run
If Echorank is unavailable or its usage limit for the period is reached, the affected rules are marked not assessable with the reason and left out of the score, never counted as failures. Each account has a monthly allowance for Echorank checks; once it is used, the checks resume the next month.