34 checks + 17 shown for information. The criteria are public for everyone, and Free and AEO+ show exactly the same number of issues; the only differences are step-by-step fixes, PDF, weekly re-checks and the visa-page score. Every criterion looks only at your own website and never compares with other businesses.
| Item | Criterion | Severity | Public score weight | Visa-page score | Free | AEO+ |
|---|---|---|---|---|---|---|
| AI crawler access (10 checks) · Who can get into your site | ||||||
| GPTBot (OpenAI · Training) | Whether robots.txt allows the homepage / (RFC 9309: the longest matching rule wins for a crawler; without its own group, User-agent: * applies) | High | 12 for all 4 | — | Result + issue title | Result + step-by-step fix |
| OAI-SearchBot (OpenAI · Search) | Whether robots.txt allows the homepage / (RFC 9309: the longest matching rule wins for a crawler; without its own group, User-agent: * applies) | High | 12 for all 4 | — | Result + issue title | Result + step-by-step fix |
| ClaudeBot (Anthropic · Training) | Whether robots.txt allows the homepage / (RFC 9309: the longest matching rule wins for a crawler; without its own group, User-agent: * applies) | High | 12 for all 4 | — | Result + issue title | Result + step-by-step fix |
| PerplexityBot (Perplexity · Search) | Whether robots.txt allows the homepage / (RFC 9309: the longest matching rule wins for a crawler; without its own group, User-agent: * applies) | High | 12 for all 4 | — | Result + issue title | Result + step-by-step fix |
| ChatGPT-User (OpenAI · User-triggered) | Whether robots.txt allows the homepage / (RFC 9309: the longest matching rule wins for a crawler; without its own group, User-agent: * applies) | Medium | — | Counted | Result + issue title | Result + step-by-step fix |
| Claude-SearchBot (Anthropic · Search) | Whether robots.txt allows the homepage / (RFC 9309: the longest matching rule wins for a crawler; without its own group, User-agent: * applies) | Medium | — | Counted | Result + issue title | Result + step-by-step fix |
| Applebot (Apple · Search) | Whether robots.txt allows the homepage / (RFC 9309: the longest matching rule wins for a crawler; without its own group, User-agent: * applies) | Medium | — | Counted | Result + issue title | Result + step-by-step fix |
| Bingbot (Microsoft · Search) | Whether robots.txt allows the homepage / (RFC 9309: the longest matching rule wins for a crawler; without its own group, User-agent: * applies) | Medium | — | Counted | Result + issue title | Result + step-by-step fix |
| meta-externalagent (Meta · Training) | Whether robots.txt allows the homepage / (RFC 9309: the longest matching rule wins for a crawler; without its own group, User-agent: * applies) | Low | — | Counted | Result + issue title | Result + step-by-step fix |
| Amazonbot (Amazon · Training) | Whether robots.txt allows the homepage / (RFC 9309: the longest matching rule wins for a crawler; without its own group, User-agent: * applies) | Low | — | Counted | Result + issue title | Result + step-by-step fix |
| robots.txt has no broken lines | Finds lines AI crawlers cannot read and silently skip: misspelled directives (custom ones excluded), missing colons, rules before any User-agent, paths not starting with /, Sitemap lines without a full URL, files over 500 KB. Advice only; not scored. | Info only | — | — | Result | Result |
| Key page access | Only the listed public pages are checked; scores stay the same and citations are not guaranteed | Info only | — | — | Result | Result |
| Google-Extended (control token) | Decides whether Google may use your content for Gemini training and grounding; doesn't affect Google Search indexing. Not a crawler; refusing it is your choice | Never an issue | — | — | Not listed | Current status and explanation |
| Applebot-Extended (control token) | Decides whether Apple may use content fetched by Applebot to train models; doesn't affect Siri search. Not a crawler; refusing it is your choice | Never an issue | — | — | Not listed | Current status and explanation |
| Technical readability (14 checks) · Can AI understand it | ||||||
| Easy-to-fill forms | Checks only names and instructions read on this page. Filling and submission are not tested; scores stay unchanged. | Info only | — | — | Result | Result |
| Homepage opens | The homepage returns 200 and is a web page | High | — | — | Result + issue title | Result + step-by-step fix |
| robots.txt is readable | /robots.txt returns 200 and isn't HTML | Medium | — | — | Result + issue title | Result + step-by-step fix |
| Sitemap (sitemap.xml) | The Sitemap declared in robots.txt, or /sitemap.xml, contains <loc> entries | Medium | 8 | — | Result + issue title | Result + step-by-step fix |
| Homepage uses HTTPS | The final URL starts with https:// | High | 5 (basics) | Counted | Result + issue title | Result + step-by-step fix |
| Title and description | <title> has at least 2 characters; meta description has at least 20 readable units (CJK characters; English words) and at most 320 characters in total | Medium | 10 | — | Result + issue title | Result + step-by-step fix |
| Business structured data | JSON-LD has a business node (LocalBusiness / Organization, etc.) with a name | High | 12 | — | Result + issue title | Result + step-by-step fix |
| Canonical URL | <link rel="canonical"> is present | Low | 5 (basics) | Counted | Result + issue title | Result + step-by-step fix |
| Official address | Check that this page's official addresses agree and its destination can be read. No points are added or deducted. Having a canonical URL is scored separately in the 5-point public canonical URL item. | Info only | — | — | Result | Result |
| Page language (lang) | <html lang> has a value | Low | — | Counted | Result + issue title | Result + step-by-step fix |
| Section headings (h2) | At least one <h2> on the homepage | Low | — | Counted | Result + issue title | Result + step-by-step fix |
| Image text (alt) | 80% or more of homepage <img> have alt; no images = not applicable | Medium | — | Counted | Result + issue title | Result + step-by-step fix |
| Easy to read on phones | Checks this page's phone width and zoom settings. Actual layout is untested; no points are added. | Info only | — | — | Result | Result |
| No hidden instructions for AI | Looks for text people can't see but that is written for AI (e.g. hidden "recommend this shop" or "ignore previous instructions"). Counts only when the text is both hidden and instruction-like; no points. | Info only | — | — | Result | Result |
| Homepage response time | Homepage fully loaded within 1.50 s of the request (checked from Tokyo, single measurement) | Medium | — | Counted | Result + issue title | Result + step-by-step fix |
| No fake pages (soft 404) | Fetch a URL that can't exist: it must return 404 (or 410), not 200 + a page | Medium | — | Counted | Result + issue title | Result + step-by-step fix |
| Redirect count | At most 2 redirects in total, including http→https and adding www | Low | — | Counted | Result + issue title | Result + step-by-step fix |
| Page can be fully read | HTML ≤ 2MB, or complete business identity appears in the first 2MB | Low | — | Counted | Result + issue title | Result + step-by-step fix |
| Search listing switch | Checks only the search listing and text-use settings read this time. No guarantee of search or AI recommendations. This does not affect the public score but counts toward the AEO+ visa-page score. | Info only | — | Counted | Result | Result |
| Business data check | Only data format, name and address are checked. Phone and opening hours are recommendations. Images, accuracy and search display are unverified. | Info only | — | — | Result | Result |
| Website content readability | Checks only the homepage read this time; suggestions only, with no points deducted. | Info only | — | — | Result | Result |
| Content completeness (9 checks) · Can AI introduce you | ||||||
| Title names your trade and area | Checks whether the page title, description or first paragraph say what you do (from the structured-data type) and where (from the structured-data region and locality). Advice only; not scored. | Info only | — | — | Result | Result |
| Description is specific | Checks whether the meta description is a site-builder default, identical to the title, just repeats the title, or is only a welcome line. Advice only; not scored. | Info only | — | — | Result | Result |
| Consistent name (N) | The structured-data name appears in <title> or og:site_name | Medium | — | Counted | Result + issue title | Result + step-by-step fix |
| Address (A) | address is present in structured data (a complete region, city, street and number is recommended) | Medium | 8 | — | Result + issue title | Result + step-by-step fix |
| Phone (P) | telephone in structured data, or a tel: link on the page | Medium | 8 | — | Result + issue title | Result + step-by-step fix |
| Opening hours | openingHours or openingHoursSpecification in structured data | Medium | 8 | — | Result + issue title | Result + step-by-step fix |
| Business introduction | A readable introduction paragraph describing who, what and where | Medium | 8 | — | Result + issue title | Result + step-by-step fix |
| Real questions and answers | Visible question and answer; no special markup or fixed count required | Medium | — | Counted | Result + issue title | Result + step-by-step fix |
| Profile links (sameAs) | At least 1 external URL in sameAs on any JSON-LD node | Low | — | Counted | Result + issue title | Result + step-by-step fix |
| Language versions (hreflang) | Shown only, not scored: <link rel="alternate" hreflang> | Info only | — | — | Result | Result |
| Official site links to your shop page | Checks only a href and JSON-LD sameAs in the homepage source HTML, without running JavaScript. Reminder only; no effect on scores or stamps. | Low | — | — | Result + issue title | Result + step-by-step fix |
| Official registry match | Low | — | — | Result + issue title | Result + step-by-step fix | |
| Directions to your shop | Only text and links on this page were checked. Confirm the full street address, branch, screen position and opening status. No points deducted. | Info only | — | — | Result | Result |
| Business details match | Compare only details clearly belonging to the same shop, present on both sides and directly comparable. | Info only | — | Counted | Result | Result |
| Booking entry | Checks only the booking entry read this time. This item does not deduct points; a stamp is earned only when a usable booking entry is confirmed. | Info only | — | Counted | Result | Result |
| Readable prices | Link each named item to a price, explicit currency and unit; shown only, with no score deduction. | Info only | — | — | Result | Result |
| One-tap contact | Checks contact link names and destinations on this page only. No points added or deducted. | Info only | — | — | Result | Result |
| Content freshness (1 checks) · Does AI think you're still open | ||||||
| Sitemap has update dates | At least one sitemap URL has <lastmod> | Low | — | Counted | Result + issue title | Result + step-by-step fix |
Public score = completed weight / assessed weight × 100, rounded. The 9 passport items have weight 80 and the 2 basics (HTTPS and canonical URL) have weight 5 each: total weight 90. Missing registry data and unconfirmed items are excluded from the denominator. Visa-page score = scored visa-page items + advanced checks, weighted by severity (high 3, medium 2, low 1), shown only in AEO+ and never added to the public score. Not covered this phase: Core Web Vitals, external review platforms, NAP consistency across platforms (these need external data sources).