AI Search Readiness Audit & Fix Plan
Build an agentic workflow that ingests your homepage HTML, scores AI-search readiness on a 100-point rubric, classifies your site Hot/Warm/Cold for answer-engine citation, and drafts the exact fixes — JSON-LD blocks, rewritten paragraphs, robots.txt additions — ready to paste in.
Pro tip
Run this quarterly on your homepage and your top 3 money pages. The first run almost always scores 30–55 — don't panic. The single highest-leverage fix is usually rewriting the first 50 words of your homepage to directly answer the query your customers actually type. Schema matters, but answer-first content is what AI engines extract.
How to use this prompt
- Pick your AI model. Choose the tab for Claude, ChatGPT, Gemini or Copilot — each variant is tuned for that model.
- Copy the full prompt. Click Copy Full Prompt to copy the text to your clipboard.
- Paste into your AI tool. Open your chosen model and paste the prompt into a new chat.
- Replace the
[placeholders]. Swap any bracketed fields for your company name, audience, product or tone. - Run and refine. Review the output. If anything is off, ask the AI to tighten tone, length or format.
Prompt Variants by Model
Claude
Claude 4.x
FRESH APR 2026
You are an AI search / AEO (Answer Engine Optimization) auditor. I am going to provide the HTML source of a web page from my business. Your job is to build a full AI-search readiness scorecard and...You are an AI search / AEO (Answer Engine Optimization) auditor. I am going to provide the HTML source of a web page from my business. Your job is to build a full AI-search readiness scorecard and the drafted fixes for every weak area.
<business_context>
Business: [NAME] — [WHAT YOU SELL OR DO IN ONE SENTENCE]
Primary customer: [e.g. "first-time homebuyers in Texas", "HR managers at 50–500 person SaaS companies"]
Geographic / market focus: [LOCAL CITY / REGION / NATIONAL / INDUSTRY]
Top 3 search queries I want to own: [QUERY 1], [QUERY 2], [QUERY 3]
Top 3 competitors AI engines cite today: [COMPETITOR 1], [COMPETITOR 2], [COMPETITOR 3]
Main conversion goal of this page: [e.g. "book a consultation", "email signup", "purchase"]
</business_context>
<scoring_rubric>
Score each dimension 0 to the max shown, with one sentence of evidence per score:
1. Structured data coverage (0–25 pts):
- Organization + WebSite JSON-LD present and valid (0–8)
- At least one page-specific schema (FAQPage, HowTo, LocalBusiness, Article, Product) present and valid (0–10)
- No duplicate microdata/JSON-LD for the same entity (0–4)
- Every required Schema.org field populated (0–3)
2. Answer-first content (0–25 pts):
- First paragraph directly answers the implied page-title question (0–10)
- Answer lands in ≤50 words, no brand-first filler (0–8)
- Contains at least one specific fact, number, or unique claim (0–7)
3. FAQ / question coverage (0–15 pts):
- 5+ visible FAQs phrased as real user queries (0–7)
- FAQPage JSON-LD present and matches visible content (0–5)
- Answers are direct, 40–90 words, factually defensible (0–3)
4. Entity clarity (0–15 pts):
- One-sentence "what does this business do?" is extractable above the fold (0–6)
- NAP (name, address, phone) or clear service area findable (0–5)
- Authority signals present (founder bios, dates, credentials, press) (0–4)
5. Crawler access (0–10 pts):
- robots.txt present, references sitemap (0–3)
- Explicit allow for GPTBot, PerplexityBot, ClaudeBot, Google-Extended, Applebot-Extended (0–5)
- No accidental noindex / nofollow on the page (0–2)
6. Citation-worthiness (0–10 pts):
- Specific numbers, dates, methodologies, or unique data an engine would quote (0–6)
- Claims are verifiable and sourced (0–4)
</scoring_rubric>
<classification>
After summing to a 0–100 score, classify as:
- Hot (75+): already citation-worthy; focus on expanding surface area
- Warm (50–74): strong foundations, specific gaps; high ROI fixes exist
- Cold (<50): foundational work needed; prioritize the single highest-leverage fix first
</classification>
Think through the page step by step before scoring. Then produce this exact output structure:
**PART 1 — Scorecard**
Markdown table: Dimension · Max · Score · Evidence (one sentence) · Priority Fix (one line).
**PART 2 — Classification & Grade**
Total score, Hot/Warm/Cold classification, and a letter grade (A / B / C / D / F) with a 2-sentence rationale.
**PART 3 — The Single Highest-Leverage Fix**
Under 120 words, written for a non-developer. Explain what to change, where, and why it matters most for my specific goal. If it requires code, provide it.
**PART 4 — Drafted Fixes (paste-ready)**
For each dimension scoring under 70% of its max, produce the concrete fix:
- If structured data is missing → full JSON-LD block, ready to paste into <head>.
- If content is weak → rewritten paragraph(s) using my brand voice.
- If FAQ is weak → 5 new Q&A pairs matching my customer's language.
- If robots.txt is wrong → corrected robots.txt content.
- If entity clarity is weak → rewritten hero section copy.
**PART 5 — Flag for Human Review**
List any items where the page source is incomplete, signals conflict, or a business decision is required. Do NOT fabricate answers.
Here is the page HTML to audit:
[PASTE YOUR HOMEPAGE HTML HERE — view source, select all, copy]
Notes: Claude 4.x handles long HTML inputs well. If the page exceeds 30k tokens, paste only the <head> + first <main> section.
ChatGPT
GPT-5
FRESH APR 2026
Act as an AI Search / AEO (Answer Engine Optimization) auditor. I''ll paste my homepage HTML. Your job: score AI-search readiness on a 100-point rubric, classify Hot/Warm/Cold, and draft every fix...Act as an AI Search / AEO (Answer Engine Optimization) auditor. I''ll paste my homepage HTML. Your job: score AI-search readiness on a 100-point rubric, classify Hot/Warm/Cold, and draft every fix ready-to-paste.
Business context:
- Name & what we do: [NAME — SENTENCE]
- Customer: [WHO]
- Market focus: [AREA]
- Top 3 queries we want to own: [Q1], [Q2], [Q3]
- Top 3 competitors AI engines cite: [C1], [C2], [C3]
- Page''s conversion goal: [GOAL]
Scoring rubric (total 100 points):
1. Structured data (0–25): Organization+WebSite valid (0–8); page-specific schema valid (0–10); no duplicate microdata (0–4); required fields populated (0–3).
2. Answer-first content (0–25): para 1 answers title''s question (0–10); ≤50 words no filler (0–8); specific facts/numbers (0–7).
3. FAQ coverage (0–15): 5+ real-user questions visible (0–7); FAQPage JSON-LD matches (0–5); answers direct 40–90 words (0–3).
4. Entity clarity (0–15): one-sentence business answer extractable (0–6); NAP visible (0–5); authority signals (0–4).
5. Crawler access (0–10): robots.txt references sitemap (0–3); allows GPTBot/PerplexityBot/ClaudeBot/Google-Extended/Applebot-Extended (0–5); no accidental noindex (0–2).
6. Citation-worthiness (0–10): specific quotable data (0–6); verifiable/sourced (0–4).
Classify: Hot (75+) / Warm (50–74) / Cold (<50).
Output sections in this order:
1. Scorecard table: Dimension · Max · Score · Evidence · Priority Fix.
2. Total, classification, letter grade A–F with 2-sentence rationale.
3. Single highest-leverage fix, under 120 words, non-developer language, include code if needed.
4. Drafted fixes for every dimension under 70% of max — full paste-ready JSON-LD blocks, rewritten paragraphs, new FAQs, corrected robots.txt, rewritten hero copy.
5. Flag human-review items — don''t fabricate answers when business input is needed.
Paste HTML now:
[HOMEPAGE HTML HERE]
Notes: GPT-5 with web browsing can fetch the URL directly — replace the paste step with: "Fetch [URL] and audit it" if browsing is enabled.
Gemini
Gemini 2.5 Pro
FRESH APR 2026
You are an expert AI Search / Answer Engine Optimization auditor. I will paste the HTML source of my homepage. Audit it against a 100-point rubric, classify it, and draft every fix.
MY BUSINESS:
-...You are an expert AI Search / Answer Engine Optimization auditor. I will paste the HTML source of my homepage. Audit it against a 100-point rubric, classify it, and draft every fix.
MY BUSINESS:
- Name & offering: [NAME — SENTENCE]
- Customer type: [WHO]
- Market: [LOCAL/NATIONAL/INDUSTRY]
- 3 queries I want to rank for in AI answers: [Q1, Q2, Q3]
- 3 competitors AI engines currently cite: [C1, C2, C3]
- This page''s conversion goal: [GOAL]
RUBRIC (100 points total):
1. Structured data coverage — 25 pts
• Organization+WebSite JSON-LD valid (8)
• Page-specific schema valid (10)
• No duplicate microdata/JSON-LD (4)
• Required fields populated (3)
2. Answer-first content — 25 pts
• Para 1 answers implied title question (10)
• ≤50 words, no filler (8)
• Specific facts/numbers (7)
3. FAQ coverage — 15 pts
• 5+ real-user questions visible (7)
• FAQPage JSON-LD matches visible content (5)
• Direct 40–90-word answers (3)
4. Entity clarity — 15 pts
• Extractable one-sentence business answer (6)
• NAP or service area visible (5)
• Authority signals present (4)
5. Crawler access — 10 pts
• robots.txt + sitemap link (3)
• AI-crawler allow-list (5)
• No accidental noindex (2)
6. Citation-worthiness — 10 pts
• Specific quotable data (6)
• Verifiable/sourced (4)
CLASSIFICATION: Hot (75+) / Warm (50–74) / Cold (<50).
OUTPUT ORDER:
1. Scorecard table — Dimension | Max | Score | Evidence | One-line Fix.
2. Total + classification + A–F grade + 2-sentence rationale.
3. Single highest-leverage fix, <120 words, non-developer language.
4. Drafted fixes for every dimension <70% of max: full JSON-LD blocks, rewritten paragraphs, new FAQs, corrected robots.txt, rewritten hero copy — all paste-ready.
5. Human-review flags where business input is required; do not fabricate.
HTML source:
[PASTE HOMEPAGE HTML]
Notes: Gemini 2.5 Pro can read URLs directly via Deep Research — for richer audits, prepend: "Use Deep Research to fetch [URL] and any subpages linked from the main nav."
Copilot
Copilot (GPT-5)
FRESH APR 2026
You''re an AI Search / AEO (Answer Engine Optimization) auditor. I''ll paste my homepage HTML. Score it on a 100-point rubric, classify Hot/Warm/Cold, and draft every fix paste-ready.
Business...You''re an AI Search / AEO (Answer Engine Optimization) auditor. I''ll paste my homepage HTML. Score it on a 100-point rubric, classify Hot/Warm/Cold, and draft every fix paste-ready.
Business context:
• Name / what we do: [SENTENCE]
• Customer: [WHO]
• Market focus: [AREA]
• Top 3 queries I want to own: [Q1, Q2, Q3]
• Top 3 competitors AI engines cite: [C1, C2, C3]
• Conversion goal of page: [GOAL]
Rubric (100 pts):
1. Structured data (25): Organization+WebSite valid (8) | page schema valid (10) | no duplicate microdata (4) | required fields populated (3)
2. Answer-first content (25): para 1 answers title (10) | ≤50 words no filler (8) | specific facts (7)
3. FAQ (15): 5+ real questions (7) | FAQPage JSON-LD matches (5) | answers 40–90 words direct (3)
4. Entity clarity (15): one-sentence answer extractable (6) | NAP visible (5) | authority signals (4)
5. Crawler access (10): robots+sitemap linked (3) | AI crawler allow-list (5) | no noindex (2)
6. Citation-worthiness (10): specific quotable data (6) | verifiable/sourced (4)
Classify: Hot 75+ / Warm 50–74 / Cold <50.
Output sections in order:
1. Scorecard table: Dimension · Max · Score · Evidence · Priority Fix
2. Total + classification + grade A–F + 2-sentence rationale
3. Highest-leverage single fix (<120 words, non-developer language, include code if needed)
4. Drafted fixes for every dimension <70% of max — paste-ready JSON-LD, rewritten paragraphs, new FAQs, corrected robots.txt, new hero copy
5. Human-review flags — don''t fabricate business decisions
HTML here:
[PASTE]
Notes: Copilot in Microsoft 365 can pull from a URL via the Edge sidebar — open the page first, then invoke Copilot and paste the prompt.
Frequently Asked Questions
What does the AI Search Readiness Audit & Fix Plan prompt do?
Build an agentic workflow that ingests your homepage HTML, scores AI-search readiness on a 100-point rubric, classifies your site Hot/Warm/Cold for answer-engine citation, and drafts the exact fixes — JSON-LD blocks, rewritten paragraphs, robots.txt additions — ready to paste in.
Which AI models is this prompt tested on?
This prompt is field-tested on Claude, ChatGPT, Gemini and Copilot. Each model has its own optimized variant above.
Do I need a paid AI account to use this prompt?
No. This prompt is written to run on the free tier of Claude, ChatGPT, Gemini and Copilot. Paid tiers simply give you longer context windows and faster responses.
Can I customize this prompt for my business?
Yes. Any text inside square brackets is a placeholder you replace with your own business details, such as company name, audience, product or tone. You can also ask the AI to adjust format, length or style after the first output.
When was this prompt last verified?
Each model variant above shows its own freshness stamp. AlignAI re-verifies every prompt at least monthly and rebuilds when a major model changes.
Don’t see what you need? tailored to your use case.