Methodology
How Answer Patch scores your site
Answer Patch uses six AI search readiness pillars and a deterministic rules engine — no AI model in the score. A free scan scores homepage pillars that actually run. Competitive visibility, robots.txt, llms.txt, live-fetch, and sitemap checks run at paid Fix Report depth. The overall score is a weighted blend of included pillars, not a ranking forecast.
Last reviewed: August 13, 2026
Pillars, checks, and scoring
A free scan scores the homepage pillars that actually run: Business Clarity, Answer Readiness, Technical Access (on-page checks), Structured Data, and Trust Signals. Competitive Visibility does not run on a free scan — the category is marked not_run and excluded from that overall score. The checks below describe the full method. Robots.txt, llms.txt, sitemap, per-crawler fetch comparison, and competitor checks run at paid Fix Report depth. Findings carry a severity (Critical, High, Medium, Low, or Info). Scored findings use Observed or Assessment confidence; Not verified means the check was not confirmable from available evidence. The checks match those listed on each feature pillar page.
| Pillar | Check | When it scores |
|---|---|---|
| Business Clarity | Title tag: is it present, long enough to carry a signal, and — for local-service sites — does it name category or city instead of a bare brand? | Scores on the free scan |
| Business Clarity | Meta description: is one present, and is the length in a range a snippet can use? | Scores on the free scan |
| Business Clarity | H1: does the page have one primary heading, not zero and not a stack of competing H1s? | Scores on the free scan |
| Business Clarity | Heading hierarchy: do heading levels step H1 → H2 → H3, or skip a level? | Scores on the free scan |
| Business Clarity | Image alt text: are selected images missing an alt attribute? Dummy values such as image1 are not this rule. | Scores on the free scan |
| Answer Readiness | Homepage opener: does the first meaningful sentence state what the business is, who it serves, and where? | Scores on the free scan |
| Answer Readiness | FAQ coverage: does the homepage show at least two visible questions with visible answers an assistant can lift? | Scores on the free scan |
| Answer Readiness | Local intent (local-service sites): does the homepage mention the target city and the services you actually offer? | Scores on the free scan |
| Answer Readiness | Service-page clarity (local-service sites): do service URLs state service, location, audience, proof, and a booking or contact step? | Scores on the free scan |
| Answer Readiness | Citability analysis in the report: passage-level scores for direct-answer openings — a report section, not a second copy of this pillar's score | Scores on the free scan |
| Technical Access | Fix Report Technical Access: crawler allow-block, sitemap, llms.txt, and live-fetch comparison | Paid Fix Report only — not scored on the free scan |
| Technical Access | robots.txt rules: are AI crawlers like GPTBot, ClaudeBot, and PerplexityBot allowed or blocked? | Paid Fix Report only — not scored on the free scan |
| Technical Access | llms.txt presence and quality: does the site publish a curated summary for AI systems? | Paid Fix Report only — not scored on the free scan |
| Technical Access | Server-side rendering: does JavaScript-only rendering hide content from non-browser fetches? | Scores on the free scan |
| Technical Access | Response codes and redirects: do crawled pages return clean 200s or break on redirect chains? | Scores on the free scan |
| Technical Access | Sitemap availability: can crawlers discover the important pages? | Paid Fix Report only — not scored on the free scan |
| Technical Access | Per-crawler fetch comparison: do different AI user agents get different responses? | Paid Fix Report only — not scored on the free scan |
| Structured Data | LocalBusiness JSON-LD (local-service sites with verified local presence): present, required fields, recommended fields | Scores on the free scan |
| Structured Data | FAQPage quality: if FAQ schema exists, do questions include acceptedAnswer? Do schema questions match visible copy? | Scores on the free scan |
| Structured Data | FAQ page expectation: if the page is an FAQ URL with visible questions, is FAQPage present? | Scores on the free scan |
| Structured Data | BreadcrumbList: inner pages should declare a trail; items need ordered positions and absolute URLs | Scores on the free scan |
| Structured Data | Absolute schema URLs: url, logo, image, and sameAs should be http(s), not relative paths | Scores on the free scan |
| Structured Data | Page-type expectations: LocalBusiness on a local homepage; Organization, WebSite, and SoftwareApplication on SaaS; Article or BlogPosting on editorial pages for SaaS, ecommerce, content, and other — not for local_service | Scores on the free scan |
| Trust Signals | Credential tokens in page text: Licensed, Certified, Accredited, DDS, MD, and the other tokens the extractor looks for — each match is one proof point | Scores on the free scan |
| Trust Signals | Numeric reviews mention: one bucket if main text matches a number plus stars, reviews, or rated — a quote with no number does not count | Scores on the free scan |
| Trust Signals | Featured-on listing links: visible directory or press links in the body, not a footer social-icon row | Scores on the free scan |
| Trust Signals | Schema sameAs (separate rule): does Organization or LocalBusiness JSON-LD include at least two profile URLs? | Scores on the free scan |
| Trust Signals | Not scored here: case studies, customer stories, years-in-business, generic testimonials without a number (Competitive Visibility); authorship; physical address (LocalBusiness in Structured Data) | Scores on the free scan |
| Competitive Visibility | Fix Report competitive comparison: six observed dimensions against the competitor URLs you add | Paid Fix Report only — not scored on the free scan |
| Competitive Visibility | Answer readiness vs. competitors: are extractable passages as easy to quote as the pages you named? | Paid Fix Report only — not scored on the free scan |
| Competitive Visibility | FAQ coverage: do competitors publish visible FAQ answers and matching FAQPage schema you lack? | Paid Fix Report only — not scored on the free scan |
| Competitive Visibility | Schema coverage: which schema.org types appear on comparable pages? | Paid Fix Report only — not scored on the free scan |
| Competitive Visibility | Proof signals: reviews, credentials, years in business, sameAs links, and customer stories | Paid Fix Report only — not scored on the free scan |
| Competitive Visibility | Content depth and structure: a clear H1 plus enough crawlable text for answer extraction | Paid Fix Report only — not scored on the free scan |
| Competitive Visibility | Technical access basics: viewport, canonical, and server-rendered HTML risk on the compared pages | Paid Fix Report only — not scored on the free scan |
Score bands
These bands are the same labels a Fix Report uses.
| Range | Label | Meaning |
|---|---|---|
| 80-100 | Strong | Most audited signals are clear, fetchable, and answer-ready. |
| 60-79 | Improving | The site has useful foundations but meaningful gaps remain. |
| 0-59 | Needs work | Important crawlability, clarity, schema, or proof signals are missing. |
How overall weights vary by business type
The paid Fix Report overall score is a weighted blend of the six pillars. On a free scan, Competitive Visibility does not run and is excluded from the overall score; remaining included pillars are renormalized. The map below is the scoring engine's weight table, not a marketing estimate. Weights in the table always sum to 100.
| Pillar | Local service | SaaS | E-commerce | Content | General |
|---|---|---|---|---|---|
| Business Clarity | 20 | 20 | 15 | 15 | 20 |
| Answer Readiness | 20 | 20 | 20 | 25 | 20 |
| Technical Access | 15 | 15 | 15 | 15 | 15 |
| Structured Data | 15 | 10 | 15 | 10 | 10 |
| Trust Signals | 20 | 15 | 15 | 20 | 20 |
| Competitive Visibility | 10 | 20 | 20 | 15 | 15 |
Sample findings
Illustrative examples of issues a free scan or paid Fix Report can flag — not from a named customer, and not a traffic or citation metric. Fix Report-only examples are labeled.
Business Clarity · High
Homepage title lacks category or location signal — brand name only
Answer Readiness · High
Homepage lacks a direct-answer opener
Technical Access · High · Fix Report only
robots.txt blocks ChatGPT-User — live citation fetch is prevented
Structured Data · High
LocalBusiness schema is missing
Trust Signals · Medium
No visible proof signals detected
Competitive Visibility · Low · Fix Report only
Two category competitors publish FAQ schema that your site lacks
Illustrative examples — not from a real scan of a named business.
Which AI crawlers the audit checks
At paid Fix Report depth, Technical Access has two crawler checks, and they are not the same list. The robots.txt table is the list the scoring engine evaluates for allow/disallow rules. A separate live-fetch comparison then requests the homepage as a smaller set of user-agents. The free scan does not run these crawler checks. We do not fetch as every crawler in the robots.txt table, and we do not operate those crawlers.
Robots.txt checks
The paid Fix Report reads robots.txt for these user-agents — the same roster the scoring engine uses. Purpose is training, fetch, or both, as the engine records it. Google-Extended is a training token; the live Google fetch uses Googlebot.
| Platform | User-agent | Purpose |
|---|---|---|
| ChatGPT / OpenAI | GPTBot | training |
| ChatGPT browsing | ChatGPT-User | fetch |
| OpenAI Search | OAI-SearchBot | fetch |
| Anthropic / Claude | ClaudeBot | training |
| Claude per-user fetch | Claude-User | fetch |
| Claude search | Claude-SearchBot | fetch |
| Anthropic (legacy) | anthropic-ai | training |
| Perplexity AI | PerplexityBot | both |
| Google Gemini / Bard | Google-Extended | training |
| Apple AI features | Applebot-Extended | training |
| Amazon Alexa / AI | Amazonbot | both |
| ByteDance / Doubao | Bytespider | both |
| Common Crawl | CCBot | training |
| Meta AI | meta-externalagent | both |
| You.com | YouBot | both |
| Cohere | cohere-ai | training |
| Turnitin AI | TurnitinBot | both |
Live fetch comparison
At paid Fix Report depth, live fetch comparison requests the homepage as these user-agents only. It is not a fetch of every crawler in the robots.txt table.
- Googlebot
- OAI-SearchBot
- ChatGPT-User
- PerplexityBot
- Claude-User
- Claude-SearchBot
Primary sources
Crawler names and purposes come from vendor documentation. Schema checks follow schema.org types. Findings reflect public pages and files available at audit time. Score recovery estimates in a report are internal score estimates, not traffic, ranking, or citation guarantees.
- ChatGPT / OpenAI crawler documentation
- ChatGPT browsing crawler documentation
- OpenAI Search crawler documentation
- Anthropic / Claude crawler documentation
- Perplexity AI crawler documentation
- Google Gemini / Bard crawler documentation
- Apple AI features crawler documentation
- Amazon Alexa / AI crawler documentation
- Common Crawl crawler documentation
- Meta AI crawler documentation
- schema.org
Frequently asked questions
How does Answer Patch score a website?
Answer Patch uses six AI search readiness pillars and a deterministic rules engine — no AI model in the score. A free scan scores homepage pillars that actually run. Competitive visibility, robots.txt, llms.txt, live-fetch, and sitemap checks run at paid Fix Report depth. The overall score is a weighted blend of included pillars, not a ranking forecast.
What does each pillar score mean?
Each pillar score runs from 0 to 100. 80-100 is Strong: Most audited signals are clear, fetchable, and answer-ready. 60-79 is Improving: The site has useful foundations but meaningful gaps remain. 0-59 is Needs work: Important crawlability, clarity, schema, or proof signals are missing.
Is the score produced by an AI model?
No. Baseline findings and pillar scores come from deterministic rules. The same site produces the same baseline findings every time. AI models are used later, in the paid Fix Report, to draft copy-ready edits — and those sections are labeled.
Which AI crawlers does the audit check?
At paid Fix Report depth, the Technical Access pillar checks robots.txt against the AI crawlers in the robots.txt table on this page — the same list the scoring engine uses. Live fetch comparison is a separate paid check: it requests the homepage as Googlebot, OAI-SearchBot, ChatGPT-User, PerplexityBot, Claude-User, and Claude-SearchBot. The free scan does not run these crawler checks. We do not fetch as every crawler in the robots.txt table, and we do not operate those crawlers.
Do the sample findings on this page come from a real business?
No. Sample findings are illustrative examples of issues a free scan or paid Fix Report can flag. Competitive Visibility and crawler findings are Fix Report-only. They are not from a named customer and they do not include invented traffic or citation metrics.
Does a higher score mean ChatGPT will cite my site?
No. Answer Patch does not control how any search engine or AI system ranks, cites, or summarizes a site, and makes no ranking, traffic, or citation promise.
How are pillar weights chosen?
Overall score weights vary by business type — local service, SaaS, ecommerce, content, or general — and always sum to 100. The paid Fix Report uses the full six-pillar map on this page. On a free scan, Competitive Visibility does not run and is excluded from the overall score; remaining included pillars are renormalized.
How long does Answer Patch keep a Fix Report?
By default, a paid Fix Report and its share link are retained for 90 days from the date of purchase. Details are on the Data Retention page.
See how your site scores
A free scan scores your homepage. Robots.txt, llms.txt, live-fetch, sitemap, and competitor checks run in the paid Fix Report.
Run a free scan