AI sees only part of you. The gaps below are why you get skipped.

Recommendation receipt

Indicative, not definitive: this is a single run of one AI engine (Claude). AI answers vary run-to-run, so absence here means you're not reliably surfaced for this question — not that you're never recommended. The tracked, multi-engine, multi-run version is paid monitoring.

The 3 fixes that matter most.

1. 1
   Your content isn't shaped the way AI answers favorOn your site

   Content is missing levers proven to lift AI visibility (stats, quotes, citations).

   Passages with statistics, quotations and cited sources are extracted into AI answers more often; lead with a direct, self-contained answer.

   Strong/causal: GEO (KDD 2024) — stats +33%, quotes +41%, cited sources +28%.

2. 2
   AI crawlers can't read most of your contentOn your site

   ~11225 characters of text — readable, but thin; more substantive content helps.

   AI crawlers don't run JavaScript; client-side-injected content is invisible to them. Server-render or pre-render key content.

   Strong: 500M+ fetch study — AI crawlers don't execute JS.

3. 3
   No visible freshness datesOn your site

   No visible freshness/last-updated dates.

   AI retrieval favors recent content; expose accurate last-updated dates.

   Expose a last-updated date (replace the date with your real one)

   <script type="application/ld+json">
   {"@context":"https://schema.org","@type":"WebPage","dateModified":"YYYY-MM-DD"}
   </script>

   Moderate: AI-cited content skews recent.

Can AI agents act on your site?

Can AI agents -- not just answer engines -- find, read, call, and buy from you? Anchored to shipping standards (AI crawlers, MCP, NLWeb, ACP/AP2).

The full Agent-Readiness audit probes your live MCP / agent-card / checkout surfaces -- part of paid monitoring.

Agent Task Replay

Can an AI agent complete a sale on miro.com?

Here is what Agent Task Replay reveals (example below). Run it on miro.com with Pro.

• +
  AI crawlers can reach you

  AI retrieval crawlers (ChatGPT, Perplexity, Google) are allowed — good.

• +
  Snippet eligible

  Indexable and snippet-eligible.

• ~
  AI can read your content

  ~11225 characters of text — readable, but thin; more substantive content helps.

  Add more substantive server-rendered text.

• ~
  Content AI loves to quote

  Content is missing levers proven to lift AI visibility (stats, quotes, citations).

  Lead with a direct, answer-first summary of what you do.

• -
  AI knows you as an entity

  No Wikidata/Wikipedia entity found — AI models may not recognize your brand as a real entity.

  Establish a Wikidata entity (and ideally a Wikipedia page) — the strongest free off-page signal that AI knows you.

• -
  Freshness signals

  No visible freshness/last-updated dates.

  Expose dateModified / last-updated dates — AI retrieval favors recent content.

• +
  Sitemap discoverable

  Sitemap declared in robots.txt.

• ~
  Structured datahygiene -- not a ranking factor

  No structured data. Minor — schema is hygiene, not a proven AI-citation driver.

• -
  /llms.txt presentinformational -- not a ranking factor

  /llms.txt is informational only — major AI engines do not currently use it (it does not affect this grade).

Competitor share-of-voice

Who does AI recommend instead?

This free grade shows your readiness diagnosis only. Measured citation share-of-voice, competitor rankings, and trends over time are part of paid monitoring -- never faked here.