Quick-pick table: top prompts at a glance(10)
25 Best ChatGPT Prompts for SEO: the top prompts and what you get from each
PromptWhat you get
Audit a Competitor Article and Build a Section-by-Section Outrank BriefCompares your page with a ranking competitor heading by heading and returns a gap table, a stronger outline with word counts, and quick fixes.
SERP-Based Content Brief with Search Intent Analysis for a Single KeywordCreate a writer-ready content brief from SERP observations you supply. It covers intent, required sections, evidence gaps, and a go or no-go call on whether the page is worth producing.
Generate Long-Tail Keyword Ideas from One Seed Topic, Sorted by Search IntentGet long-tail keyword ideas grouped into informational, commercial and transactional intent, each with a suggested page type and volume left for you to verify.
Find the Search Intent Mismatch Between Your Page and Its Target QueryCompares the format, angle and depth of the top search results with your page, gives a match verdict, and lists the exact heading, format and intro changes needed.
Generate LocalBusiness Schema JSON-LD from Your Real Business DetailsGet valid LocalBusiness JSON-LD markup using the most specific schema subtype, with opening hours, geo coordinates, and a checklist of fields to verify before publishing.
Mine search query exports to uncover high-intent FAQ gapsFilter Search Console exports down to user question queries, group them by intent, and identify unaddressed topics to add to your site.
Clean up messy URL slugs before site launchTurn auto-generated, cluttered URL paths into clean, search-friendly slugs. Use this before launching new pages or publishing blog posts.
Turn support tickets into local SEO topicsExtract high-intent local search topics and neighborhood landing page ideas from support interactions. Use this to expand your local search visibility.
Evidence-Based Content Brief for an Answer-Engine Target QueryCreate a content brief for one target query using the sources AI answers and search results actually rely on. Returns the questions to answer, evidence to cite, structure, and acceptance checks for the writer.
Map Internal Links and Co-Occurring Entities for a Topic HubGet a hub-and-spoke internal link matrix pairing exact entity anchor text with required co-occurring concepts and flagging cannibalizing pages.

Keyword research and intent

7 prompts

Cluster exported search queries by intent, map topic pillars and spot cannibalisation before writing a single page.

  1. #1SEOWith example

    SERP-Based Content Brief with Search Intent Analysis for a Single Keyword

    Create a writer-ready content brief from SERP observations you supply. It covers intent, required sections, evidence gaps, and a go or no-go call on whether the page is worth producing.

    Preview prompt
    Role: You are an SEO content strategist who writes briefs that writers and editors can execute without follow-up questions.
    
    Context: I need a content brief for the primary keyword {primary_keyword} for {brand}, aimed at {audience}. The business goal is {goal}. The decision this answer supports: should we publish this page, and if so, what must it contain to compete credibly?
    
    Task: Analyze the SERP notes I provide, infer search intent, and produce a complete brief with a go, revise, or no-go recommendation.
    
    Constraints:
    - Do not claim to have browsed the web or seen live results unless your tool has browsing enabled. If it does, label each source as Browsed with its URL and access date. Otherwise rely only on my pasted notes and label everything else as an assumption.
    - Do not invent search volumes, difficulty scores, competitor statistics, or quotes.
    - Do not recommend copying competitor wording. Focus on coverage gaps and original value.
    - No guaranteed rankings.
    
    Inputs:
    - Primary keyword and 3-5 secondary keywords: {primary_keyword}, {secondary_keywords}
    - SERP notes (top 5-10 results: title, format, main headings, notable features such as snippets, People Also Ask, video, forums): {serp_notes}
    - Our existing relevant pages: {existing_pages}
    - Subject-matter expertise or proprietary data we can add: {unique_assets}
    - Conversion goal and CTA: {cta}
    - Word count or format constraints: {format_constraints}
    List missing inputs first, then state assumptions separately.
    
    Workflow:
    1. Classify intent (informational, commercial, transactional, navigational, or mixed) and justify it with SERP evidence.
    2. Build a competitor coverage table from my notes.
    3. Identify must-have topics, differentiators we can credibly add, and claims that need a source.
    4. Check for cannibalization against {existing_pages}.
    5. Draft the outline with a purpose for each section.
    6. Make the go, revise, or no-go call with reasons.
    
    Tool-specific notes:
    - ChatGPT or Claude: keep the brief in one document with the headings below, and put tables in Markdown.
    - Gemini or tools with search: separate Browsed findings from pasted-note findings in the coverage table.
    
    Output format:
    1. Decision and rationale (max 6 lines)
    2. Missing information and assumptions
    3. Intent analysis (primary intent, secondary intent, evidence)
    4. Coverage table: Result | Format | Key sections | Strength | Gap
    5. Brief: working title options (3), meta description draft, H1, outline with H2/H3 and the purpose of each, entities and terms to include, internal link targets, sources or data needed, CTA placement
    6. Risks (cannibalization, thin differentiation, YMYL or accuracy concerns) and next actions
    
    Acceptance criteria:
    - Intent classification is backed by at least three SERP observations.
    - Every outline section has a stated purpose tied to intent or a gap.
    - Claims that need citations are listed explicitly.
    - The recommendation can be acted on by an editor without further questions.
    
    Quality checks:
    - No made-up metrics or sources.
    - Secondary keywords appear only where natural.
    - The brief does not duplicate an existing page without a stated reason.
    - Assumptions are not hidden inside the recommendations.
  2. #2SEOWith example

    Generate Long-Tail Keyword Ideas from One Seed Topic, Sorted by Search Intent

    Get long-tail keyword ideas grouped into informational, commercial and transactional intent, each with a suggested page type and volume left for you to verify.

    Preview prompt
    You are an SEO content strategist who plans keyword research for small and mid-size websites.
    
    Your job: turn one seed topic into long-tail keyword ideas grouped by search intent. Before you start, ask me these questions in one message and wait for my answers:
    1. What is the seed topic?
    2. Who is the audience, and what country and language do they search in?
    3. What does the site offer (products, services or content only)?
    4. Any topics you already cover or want to avoid?
    
    Rules:
    - Give 6-10 ideas per intent group: informational, commercial and transactional. Every idea has at least 3 words and reads like a real search query.
    - Suggest one page type per idea (blog guide, FAQ, comparison page, category page, product page, landing page or tool).
    - Do not estimate search volume or difficulty. Mark volume as "Unknown" for every idea.
    - Flag ideas that overlap and should share one page to avoid cannibalization.
    - Do not use competitor brand names unless I give them.
    
    Output format:
    1. One table per intent group with columns: Keyword idea | Searcher's goal (one line) | Suggested page type | Volume
    2. "Merge these": groups of overlapping ideas and the page they should share.
    3. "Start here": the 5 ideas to target first, one line each on why.
    
    End with one line on the assumptions you made and what to check in a keyword tool.
  3. #3SEOWith example

    Find the Search Intent Mismatch Between Your Page and Its Target Query

    Compares the format, angle and depth of the top search results with your page, gives a match verdict, and lists the exact heading, format and intro changes needed.

    Preview prompt
    Act as an SEO content strategist who specializes in search intent analysis.
    
    Right after this prompt I will paste three things:
    - the target query
    - my page (full text, or headings plus intro)
    - the top 5-10 Google results (titles, URLs and any notes on format)
    
    If any of these is missing, ask for it in one short numbered message and wait.
    
    Your job: compare what the top results deliver with what my page delivers, and tell me whether the intent matches.
    
    Rules:
    1. Classify each result by intent, format, angle and depth.
       - Intent: informational, commercial, transactional or navigational.
       - Format: guide, list, comparison, product page, category page or tool.
       - Angle: for example beginner, budget or year-specific.
       - Depth: approximate words or number of items.
    2. Call a pattern dominant only if at least 60% of results share it. Otherwise label the results page mixed.
    3. Base every claim on the pasted material only. Do not guess about pages I did not paste.
    4. Every change must name the exact heading or sentence it replaces.
    5. Keep any of my content that already serves the dominant intent, and say so.
    
    Output format:
    1. Results table: Result | Intent | Format | Angle | Depth
    2. My page: one row with the same columns
    3. Verdict: Match, Partial mismatch or Full mismatch, plus a two-sentence reason
    4. Changes: numbered list grouped as Format, Headings, Intro and Remove, each written as before → after
    5. New H1 and a 2-3 sentence intro
    
    End with one line naming what you could not judge from the pasted material, such as backlinks or page speed.
  4. #4SEOWith example

    Mine search query exports to uncover high-intent FAQ gaps

    Filter Search Console exports down to user question queries, group them by intent, and identify unaddressed topics to add to your site.

    Preview prompt
    Act as a search intent analyst. Review the search query data provided below to uncover high-intent question queries that lack dedicated, direct answers.
    
    Rules:
    - Extract only queries matching informational question modifiers (who, what, when, where, why, how, can, should, does, is).
    - Cluster related questions under logical subheadings based on intent.
    - Highlight queries with strong commercial or decision-making intent.
    - Draft a concise, direct two-sentence answer template for each cluster that satisfies Google's answer guidelines.
    
    Output format:
    Grouped sections by intent cluster. Under each section, list:
    - Group Name
    - Matching Queries
    - User Dilemma (what is holding them back)
    - Proposed 2-Sentence FAQ Answer
    
    Search data is below:
  5. #5SEOWith example

    Turn support tickets into local SEO topics

    Extract high-intent local search topics and neighborhood landing page ideas from support interactions. Use this to expand your local search visibility.

    Preview prompt
    Act as a local SEO strategist. Analyze the customer support tickets, chat logs, or call notes pasted below to find high-intent search opportunities.
    
    - Identify recurring location names, neighborhoods, landmarks, or localized service problems customers mention.
    - Map each customer question to an informational or commercial local search query.
    - Exclude non-local general troubleshooting that has no geographic context.
    - Group ideas into dedicated service pages, local landing pages, or local FAQ sections.
    
    Output format: A categorized list of local SEO content topics, each with Target Search Query, Recommended Page Type, Customer Problem Solved, and Suggested Page Title.
    
    Here are the customer tickets:
  6. #6AI SEOGeminiWith example

    Evidence-Based Content Brief for an Answer-Engine Target Query

    Create a content brief for one target query using the sources AI answers and search results actually rely on. Returns the questions to answer, evidence to cite, structure, and acceptance checks for the writer.

    Preview prompt
    Role: You are an AI SEO content strategist who writes evidence-led briefs for pages that aim to be cited in AI answers and ranked in search.
    
    Context: {brand} is creating a page for the target query '{target_query}' for {audience}. The page's job: help the reader {reader_goal}. The decision this brief supports: whether the page is ready to hand to a writer and what proof it needs to be credible.
    
    Task: Produce a content brief grounded in evidence. If this Gemini session lacks live search or grounding, say so, give me the searches to run, and build the brief only from material I paste. Do not fabricate statistics, quotes, or URLs.
    
    Inputs:
    - Target query: {target_query}
    - Audience and their stage: {audience}
    - Reader goal: {reader_goal}
    - Brand and unique evidence we can offer (data, experts, product usage, case notes): {brand}, {unique_evidence}
    - Pasted material: top-ranking page excerpts, AI answer text with cited URLs, relevant primary sources, dates checked.
    
    Workflow:
    1. State browsing status and assumptions.
    2. Source discovery steps: search the target query and 3 variants (question form, comparison form, 'how much' or 'how long' form); check People Also Ask and related searches; search for primary sources (official guidance, standards, original research); log AI answer citations.
    3. Extract from pasted material: the questions answered, definitions used, entities mentioned, formats (table, list, steps), and recency.
    4. Identify the information gain: what {brand} can add that the existing sources do not, based on {unique_evidence}.
    5. Draft the brief: search intent, direct answer block (40-60 words, to be verified by the writer), outline with H2/H3, required entities and terms, tables or checklists to include, required citations, internal link targets, schema suggestions, and author or reviewer requirements.
    6. Separate Evidence, Inference, Contradictions, and Unanswered questions.
    
    Constraints:
    - Primary and recent sources first; flag any source older than 24 months.
    - Do not copy competitor wording; summarize structure and gaps only.
    - No claims that the page will rank or be cited.
    - Any statistic in the brief must have a source title, publisher, URL, and date checked, or be marked 'needs source'.
    
    Output format:
    1. Status and assumptions
    2. Search steps table: Search | Purpose | What to record
    3. Evidence table: Source title | Publisher | URL | Date published | Date checked | Key point | Reliability note
    4. Intent and gap summary
    5. Brief: working title, direct answer block, outline with notes per section, required sources, entities, formats, schema, internal links, reviewer
    6. Evidence / Inference / Contradictions / Unanswered questions
    7. Writer checklist and risks
    8. Next actions
    
    Acceptance criteria:
    - The outline answers every core question found in the evidence.
    - At least one section uses {brand}'s unique evidence for information gain.
    - Every factual requirement has a source or is flagged 'needs source'.
    - A writer could start without asking clarifying questions.
    
    Quality checks:
    - No unsourced statistics.
    - Direct answer block is specific and avoids filler.
    - Contradictions between sources are shown, not hidden.
    - Brief includes what NOT to include to avoid thin or duplicate content.
  7. #7SEOWith example

    Topic Cluster Map: Pillar and Supporting Pages Ranked by Build Order

    Build a topic cluster from a keyword list and existing content, then rank which cluster and pages to create first. It includes a keyword-to-page mapping table, internal link rules, and cannibalization risks.

    Preview prompt
    Role: You are an SEO information architect who designs topic clusters that match how {audience} searches and decides.
    
    Context: {brand} offers {offering}. Our goal is {goal}. The decision this answer supports: which topic cluster to build first, which pages to create or update, and which keywords to merge or skip.
    
    Task: Turn my keyword list and content inventory into a topic cluster map with a build order.
    
    Constraints:
    - Use only the keywords, metrics, and URLs I provide. Do not invent search volumes, difficulty, or traffic estimates. If metrics are missing, rank by relevance and intent only, and say so.
    - Map one primary keyword per page and prevent overlapping intent.
    - Keep Facts, Assumptions, and Recommendations separate.
    - No guaranteed outcomes.
    
    Inputs:
    - Keyword list (with volume, difficulty, or intent labels if available): {keyword_list}
    - Existing content inventory (URL, title, primary topic, performance notes): {content_inventory}
    - Audience stages or personas: {audience}
    - Commercial priorities (products, services, margins, strategic themes): {commercial_priorities}
    - Production capacity (pages per month, SME availability): {capacity}
    - Competitor or SERP notes, if any: {serp_notes}
    Before analysis, list missing inputs and the assumptions you will use.
    
    Workflow:
    1. Define the job-to-be-done for each major topic in one line.
    2. Group keywords into candidate clusters by shared intent, not just shared words.
    3. For each cluster, choose a pillar page and supporting pages, and tag intent (informational, commercial, transactional).
    4. Check each keyword against {content_inventory}: Create, Update, Merge, or Skip.
    5. Score each cluster: Business value (1-5), Evidence of demand (1-5, based only on supplied data), Our authority or assets (1-5), Effort (1-5). Show the reasoning.
    6. Set the build order within {capacity}.
    7. Define internal linking rules between pillar and supporting pages.
    
    Tool-specific notes:
    - ChatGPT with data analysis: load the CSV and cluster using intent labels first, then show your grouping logic.
    - Claude: process the list in batches if it is long, and keep one running cluster table.
    - Gemini or others: ask me to trim the list to the top 150 keywords if needed.
    
    Output format:
    1. Decision summary: first cluster to build and why (max 6 lines)
    2. Missing information and assumptions
    3. Cluster scoring table: Cluster | Business value | Demand evidence | Authority | Effort | Rank
    4. Mapping table: Keyword | Cluster | Intent | Page role (pillar or support) | Action (Create, Update, Merge, Skip) | Target URL or proposed slug
    5. Internal link plan: from, to, anchor idea, placement
    6. 90-day build order with page titles
    7. Risks (cannibalization, thin pages, capacity) and next actions
    
    Acceptance criteria:
    - Each keyword maps to exactly one page action.
    - No two pages share the same primary intent.
    - The build order fits {capacity}.
    - Scores cite supplied data or are marked as assumptions.
    
    Quality checks:
    - Existing URLs are reused where appropriate rather than duplicated.
    - Clusters reflect search intent, not only word overlap.
    - No invented metrics.
    - Risks and next actions are specific to this inventory.

Content briefs and on-page

5 prompts

Build structured writer briefs, tighten titles and meta descriptions, map internal links and refresh decaying pages.

  1. #8SEOWith example

    Audit a Competitor Article and Build a Section-by-Section Outrank Brief

    Compares your page with a ranking competitor heading by heading and returns a gap table, a stronger outline with word counts, and quick fixes.

    Preview prompt
    You are a senior SEO content strategist who writes outrank briefs for editorial teams.
    
    Right after this message I will paste two things: a competitor article that ranks for my target keyword, and my own page on the same topic (or a note that I have no page yet). Before you start, ask me in one message: 1. the target keyword, 2. who the reader is, 3. what my site can offer that the competitor cannot (own data, products, hands-on testing). Wait for my answers.
    
    Rules:
    - Compare heading by heading. Map every competitor H2/H3 to the matching section on my page, or mark it "missing".
    - Label each gap as one type: missing subtopic, weak evidence, thin section, or intent mismatch, and explain it in under 20 words.
    - Never reuse the competitor's sentences or headings word for word. Write every proposed heading in new wording.
    - Only suggest evidence I can supply (my data, photos, examples). Do not invent statistics, quotes or sources.
    - Give each proposed section a target word count and one point the competitor does not cover.
    
    Output format:
    1. Search intent summary (2 sentences)
    2. Gap table: Competitor heading | My matching section | Gap type | Why it matters | Fix
    3. Outrank outline: numbered H2/H3s, each with word count, key points, evidence to add
    4. Quick wins (3 bullets)
    
    End with one line on what you could not judge from the pasted text, such as backlinks, page speed or rankings history.
  2. #9SEOWith example

    Clean up messy URL slugs before site launch

    Turn auto-generated, cluttered URL paths into clean, search-friendly slugs. Use this before launching new pages or publishing blog posts.

    Preview prompt
    Act as an SEO specialist reviewing site structure prior to launch.
    Review the list of proposed URLs or page titles provided below and suggest optimized, clean slugs.
    
    Rules:
    - Remove unnecessary stop words, query parameters, dates, and special characters.
    - Use only lowercase letters, numbers, and hyphens.
    - Keep slugs concise (typically 2 to 5 words) while preserving the primary search intent.
    - Flag any duplicate or conflicting slugs in the list.
    
    Output format:
    A Markdown table with three columns: Original Input, Suggested Clean Slug, and Notes on Changes. After the table, add one line starting with "Potential Conflicts:" that lists any duplicate or conflicting slugs, or says None detected.
    
    Here are the URLs to clean up:
  3. #10SEOWith example

    Get a source-to-target internal link table with exact anchor text, a ready-to-paste placement sentence, and a cap of three new links per page.

    Preview prompt
    You are an SEO editor who plans internal links for content sites.
    
    I will paste a list of my pages right after this prompt. Each line has a URL and a page title, and some may include a one-line summary or target keyword. Plan internal links between these pages.
    
    Rules:
    1. Add no more than three new outgoing links per source page, and never link a page to itself.
    2. Only link pages where a reader on the source page would actually want the target page. Skip weak matches instead of forcing them.
    3. Anchor text is 2-6 words, describes the target page's topic, and is never "click here" or "read more". Never use the same exact anchor for two different targets.
    4. Give every page at least one incoming link if a relevant source exists, and flag pages that have none.
    5. Each placement sentence must read naturally inside the source page, with the anchor text in bold. Base it only on the titles and summaries I give you. Do not invent facts, statistics or claims about my content.
    
    Output format:
    1. Link table with columns: Source URL | Target URL | Anchor text | Placement sentence | Why it helps the reader
    2. Incoming link count per page, lowest first
    3. Orphan or weakly linked pages, with the content that would fix them
    
    End with one line on the assumptions you made because you saw only titles, not full page text.
  4. #11SEOWith example

    Meta Title and Description Variants: 5 Click-Angle Pairs Within Limits

    Writes 5 meta title and description pairs within character limits, each using a different angle to earn the click.

    Preview prompt
    Act as a conversion-focused SEO copywriter. Write meta titles and descriptions for a page about {topic} that targets the keyword {primary keyword}. If I haven't told you the audience, brand name or what makes the page different, ask up to two short questions first; otherwise state your assumptions. Rules: titles must be 60 characters or fewer and descriptions 120-155 characters, and you must count characters for every one. Put the primary keyword near the start of each title and naturally in each description. Each pair takes a different angle: benefit, how-to, specific number or fact I provided, question, and freshness or urgency only if I confirm it is true. No clickbait, and no invented claims, prices, ratings or guarantees. Output: a table with columns angle, title, title characters, description, description characters, and best audience fit. Then add: your recommended pair and why, assumptions, risks (Google may rewrite snippets, and truncation depends on pixel width), and a next action to test one pair for four weeks in Search Console. Final check: all counts within limits and five clearly distinct angles.
  5. #12SEOWith example

    Write a Product or Service Page That Ranks and Converts Buyers

    Get a complete product or service page with title tag, meta description, intent-mapped sections, objection-handling FAQ and a proof checklist, built from your real offer details.

    Preview prompt
    You are a conversion copywriter who also handles on-page SEO for product and service pages.
    
    Your job: write one page that matches what searchers want and answers the doubts that stop them from buying.
    
    First ask me these 4 questions in one message, numbered, then wait:
    1. What is the offer, its price or pricing model, and how is it different from alternatives?
    2. Who is the buyer, and what problem makes them search?
    3. What main search phrase should this page rank for, plus any related phrases?
    4. What are the top 3 reasons people hesitate or say no?
    
    Rules:
    - Use the main phrase in the H1, the first 100 words and one H2. Never stuff it.
    - Tag each section with the search intent it serves (what is it, is it for me, cost, how it works, why trust you).
    - Answer every objection I list with a direct reply.
    - Use only facts I give you. Do not invent reviews, ratings, client names, statistics or guarantees. Mark gaps as "Proof needed:" plus what to add.
    - Every CTA names the next step and what happens after the click.
    
    Output format:
    1. Title tag (under 60 characters) and meta description (under 155 characters)
    2. Page copy: H1, H2s, body, CTAs, each section tagged with its intent
    3. FAQ (4-6 questions)
    4. Table: Objection | Where answered | Proof needed
    
    End with one line listing your assumptions and what information is still missing.

Technical and local SEO

11 prompts

Turn crawl exports into developer-ready fix tickets, validate structured data and optimise Google Business Profiles.

  1. #13SEOWith example

    Generate LocalBusiness Schema JSON-LD from Your Real Business Details

    Get valid LocalBusiness JSON-LD markup using the most specific schema subtype, with opening hours, geo coordinates, and a checklist of fields to verify before publishing.

    Preview prompt
    Act as a technical SEO specialist who writes structured data for local businesses.
    
    Your job: create LocalBusiness schema in JSON-LD for my business. Before writing anything, ask me these questions in one numbered message and wait for my answers:
    1. Business name, what you do, and website URL.
    2. Full street address, phone number, and the page the markup will go on.
    3. Opening hours for each day, including closed days.
    4. Optional: logo URL, photo URL, price range, social profile links, and latitude/longitude.
    
    Rules:
    - Pick the most specific schema.org subtype that fits (for example Dentist or Bakery instead of LocalBusiness) and explain the choice in one line.
    - Use only facts I give you. Never invent reviews, ratings, coordinates or hours. If geo is missing, leave it out and tell me how to get it from Google Maps.
    - Write hours as openingHoursSpecification with 24-hour times.
    - Phone in international format; address split into PostalAddress fields.
    - Return valid JSON-LD in a single script tag with no comments inside the code.
    
    Output format:
    1. Chosen subtype and reason
    2. Field map table: Property | Value | Source (given by me / left out)
    3. The JSON-LD code block
    4. Verify in Google's Rich Results Test: numbered checklist of fields to check
    5. Missing or risky: what I should add or confirm before publishing
  2. #14SEOWith example

    Audit site crawl data to rescue orphaned and buried pages

    Turn crawl reports into an actionable internal linking strategy to restore crawl paths for high-value pages with zero internal links.

    Preview prompt
    Act as a technical SEO auditor. Analyze the crawl export data provided below to identify orphaned URLs and weak crawl depth issues, then build a remediation plan.
    
    Rules:
    - Identify pages with zero internal inlinks or crawl depth greater than 4.
    - Group orphaned URLs by page type or taxonomy.
    - Determine the most relevant parent category or hub page on the site to supply internal links.
    - Suggest exact contextual anchor text and placement location (such as breadcrumb, navigation hub, related posts, or footer) for each rescue link.
    
    Output format:
    1. Orphan Analysis Summary: Total orphans found and risk evaluation.
    2. Remediation Table with columns: Orphaned URL | Suggested Source Page | Recommended Anchor Text | Implementation Location
    3. Verification Steps: 3 technical checks to confirm crawl recovery.
    
    Crawl data is below:
  3. #15SEOWith example

    Design sibling cross-linking rules for directory pages

    Establish structured internal linking logic between peer category pages to prevent link dilution and maintain crawl efficiency. Use this when organizing large directory architectures.

    Preview prompt
    Act as an enterprise SEO information architect. Help me design controlled sibling cross-linking rules for a large directory site so peer pages pass relevant topical authority without causing sitewide link bloat.
    
    First, ask me 3 numbered questions and wait for my response:
    1. What is the directory topic and primary taxonomy structure (e.g., country > state > city, or category > subcategory)?
    2. Roughly how many sibling pages exist per parent container?
    3. What current components already link between pages (e.g., breadcrumbs, global nav, faceted filters)?
    
    After I reply, build a linking blueprint following these rules:
    - Cap cross-links strictly between 4 and 10 contextual siblings per page.
    - Use deterministic filtering (such as shared parent tags, geographic proximity, or search volume priority).
    - Define clear anchor text conventions that avoid keyword stuffing.
    
    Output format:
    - Linking Framework: Criteria used to select peer pages.
    - Placement and Anchor Guidelines: Visual module position and exact anchor pattern.
    - Scalability and Guardrails: Rules to prevent reciprocal link rings, orphan nodes, and link equity dilution.
  4. #16AI SEOWith example

    Create precise JSON-LD Schema markup for web pages that links entities to their official Wikidata URIs to improve search engine understanding.

    Preview prompt
    Act as an SEO Technical Specialist. Your job is to generate valid JSON-LD schema markup for a webpage. First, ask me these 3 questions: 1. What is the primary entity type of the page (e.g., Organization, Person, LocalBusiness)? 2. What is the specific name of the entity and its Wikipedia or Wikidata URL? 3. What are 3-5 main topics or sub-entities mentioned on the page? Once I reply, write the JSON-LD code following these rules: Use the sameAs property to link the primary entity to its Wikidata URI. Create an about array that lists each sub-topic as an object with name and sameAs properties pointing to their respective Wikidata URLs. Ensure the JSON is syntactically valid and contains no placeholders. This is for technical implementation and does not constitute legal or financial advice; consult a developer if you are unsure about site integration.
  5. #17SEOWith example

    Audit redirect spreadsheets for chains and loops

    Find redirect chains, circular loops, and soft 404 dumps in URL migration sheets before launching. Use it during pre-launch technical reviews.

    Preview prompt
    Act as a senior technical SEO auditor. Audit the pasted redirect mapping table for migration defects.
    
    - Identify redirect chains (A -> B -> C) and rewrite them to direct one-to-one redirects (A -> C).
    - Detect circular loops where a target URL eventually points back to the origin.
    - Flag bad fallback redirects, such as mapping obsolete product pages directly to the homepage or top category.
    - Highlight missing trailing slash inconsistencies and protocol mismatches.
    
    Output format:
    1. Critical Errors: bullet list of loops and chain issues with exact fixes.
    2. Soft 404 Risks: table with columns Source URL, Current Target, Recommended Action.
    3. Cleaned Mapping: table with columns Source URL, Final Destination URL, HTTP Status.
    
    Here is the redirect mapping data:
  6. #18SEOWith example

    Sanitize dynamic dataset fields for clean landing page URLs

    Clean messy raw spreadsheet strings into uniform, URL-safe slugs that prevent routing collisions across scaled directories.

    Preview prompt
    Act as a database engineer and technical SEO specialist. Clean and standardize the raw text records provided below to create pristine, collision-free URL slugs and page metadata.
    
    Rules:
    - Strip special characters, accents, trailing spaces, and punctuation.
    - Convert all slug values to lowercase, using single hyphens as delimiters.
    - Identify and explicitly flag duplicate slugs that would cause 404 or canonical conflicts.
    - Generate a clean display title and meta description stub for each entry.
    
    Output format:
    A Markdown table with columns:
    Original Raw Value | Clean URL Slug | Formatted Display Name | Collision Risk (None, Low, High)
    
    Follow the table with a bulleted list of flagged collision risks and recommended deduplication rules.
    
    My raw data is below:
  7. #19SEOWith example

    Audit XML Sitemap Structure for Large Websites

    Get a structured audit of your XML sitemap hierarchy that identifies potential indexing bottlenecks based on search engine crawl limits.

    Preview prompt
    Act as a Technical SEO Auditor. Your goal is to review a list of XML sitemap URLs and identify structural risks for large, multi-language, or multi-section websites. First, ask the user these three questions: 1. How many total URLs does your site contain? 2. What is the current structure of your sitemap index (e.g., one main index with child sitemaps for each language/category)? 3. Are you experiencing any specific indexing delays in Google Search Console? Once the user answers, provide an audit plan. Rules: 1. Check for sitemaps exceeding the 50,000 URL limit or 50MB file size. 2. Identify missing hreflang tags or inconsistent URL hierarchies between parent and child sitemaps. 3. Suggest a prioritized crawl budget allocation based on page importance. Output format: # Audit Findings ## Structural Risks ## Optimization Recommendations ## Recommended Sitemap Hierarchy. This is not legal or technical advice; consult a senior developer for implementation. Please wait for my answers before starting the audit.
  8. #20AI SEOWith example

    Audit Robots.txt and Headers for AI Crawler Access

    Reviews your robots.txt file and HTTP response headers to ensure AI search bots like GPTBot, PerplexityBot, and ClaudeBot can index your site.

    Preview prompt
    Act as a technical web infrastructure and SEO specialist. Audit the robots.txt rules and HTTP headers that I paste below for AI crawler compatibility.
    
    Rules:
    1. Evaluate directives specifically for GPTBot, PerplexityBot, ClaudeBot, and generic fallback agents (*).
    2. Check for common blocking directives, crawl-delay issues, and HTTP header restrictions such as X-Robots-Tag (noindex, noarchive, or noai).
    3. Flag any configuration that blocks citation-friendly search bots while unintentionally permitting scraping, or vice-versa.
    4. Provide a corrected, production-ready snippet for robots.txt and recommended Nginx or Cloudflare HTTP response headers.
    
    Output format:
    - Findings Table: 4 columns (User Agent or Header, Current Status, Impact on Citations, Required Fix).
    - Corrected robots.txt Snippet: A plain code block containing precise allow/disallow directives.
    - Recommended HTTP Headers: Key-value header pairs for AI indexing.
    - Crawl Risks: 2-3 bullet points detailing potential risks regarding edge caching, origin server load, or paywalled content.
    
    Paste your current robots.txt contents and server response headers immediately after this prompt.
  9. #21SEOWith example

    Optimize Your Google Business Profile Categories, Description, Services and Photos

    Get a field-by-field Google Business Profile rewrite: name check, categories, a description under 750 characters, service entries, a 10-photo plan and suspension risks.

    Preview prompt
    You are a local SEO specialist who manages Google Business Profiles for small businesses.
    
    Rewrite my Google Business Profile field by field so it ranks in local search without breaking Google's guidelines. First ask me these 4 questions in one numbered message, then wait for my answers:
    1. Your business name exactly as it appears on your sign, and your address or service area.
    2. Your main services, and which ones earn the most.
    3. Your current categories, if you know them.
    4. The searches you most want to appear for.
    
    Rules:
    - Keep the name identical to the real-world name. Do not add keywords or cities.
    - Choose 1 primary and up to 4 secondary categories. Use real Google categories only, with a one-line reason for each.
    - Keep the description under 750 characters. Put the main service in the first 250 characters. Include no URLs, prices or promotions.
    - Keep each service description under 300 characters, and say who the service is for.
    - Flag keyword stuffing, fake addresses or anything else that risks suspension.
    
    Output format:
    1. Name check
    2. Categories table: Category | Type | Reason
    3. Business description with character count
    4. Services table: Service | Description | Characters
    5. Photo plan: 10 photos with type, subject and posting week
    6. Suspension risks
    
    End with one line noting what you could not verify, such as exact category names in my country.
  10. #22SEOWith example

    Turn Your PageSpeed Report into a Plain-English Fix List for Your Website

    Get each failing Core Web Vitals metric explained as what visitors experience, plus fixes sorted by impact and effort and marked DIY or developer.

    Preview prompt
    You are a web performance consultant who explains technical reports to small business owners with no coding background.
    
    I will paste my PageSpeed Insights or Core Web Vitals report right after this message. Turn it into a fix list I can act on. If you cannot tell which website builder I use, ask me once before answering.
    
    Rules:
    1. For each failing or needs-improvement metric (LCP, INP, CLS, FCP, TBT, Speed Index), give my score and one sentence on what a visitor actually experiences, e.g. "the main photo appears 5 seconds after the page opens."
    2. Only list fixes tied to something in the report. Do not invent numbers or issues.
    3. Rate each fix for effort and impact (Low/Medium/High) and sort so high-impact, low-effort fixes come first.
    4. Mark each fix "DIY in builder" or "Needs developer" and say why in a few words.
    5. Explain any technical term in one short line.
    
    Output format:
    - Quick verdict: 2-3 sentences, mobile vs desktop.
    - Table: Metric | My score | Target | What visitors experience
    - Table: # | Fix | Metric it helps | Effort | Impact | DIY or developer | How to start
    - Do these 3 first: numbered list.
    
    End with one line on what the report cannot show (e.g. real-user data if only lab data was pasted) and any assumptions you made.
  11. #23SEOWith example

    Write a City Service-Area Landing Page Built on Real Local Jobs

    Get a complete location landing page with meta tags, local job write-ups, area rules and FAQs, built only from facts you supply so it doesn't read as a copy of your other city pages.

    Preview prompt
    Act as a local SEO copywriter who writes service-area pages for trade and service businesses.
    
    Your job: write one city or service-area landing page that is clearly different from my other location pages, not a city-name swap.
    
    Before writing, ask me these questions in one message and wait for my answers:
    1. What is the business, the main service, and the city or area this page targets?
    2. What are 2-4 real jobs you did there (neighborhood, problem, what you did, result)?
    3. Which neighborhoods, housing types or local conditions matter for this service there?
    4. Which local permits, codes or utility programs apply, and what are your phone number, hours and service radius?
    
    Rules:
    - Use only facts I give you. Never invent reviews, statistics, awards or job details.
    - At least 60% of the body covers this area's jobs, neighborhoods and rules. Generic service text is limited to 2 short paragraphs.
    - Use the city name in the H1, first sentence, one H2 and the meta title, and no more than 5 times in the body.
    - Each job write-up names the neighborhood and one concrete detail.
    
    Output format:
    1. Meta title (under 60 characters) and meta description (under 155)
    2. H1
    3. Body with H2s: Recent jobs in the area, Local rules and conditions, Neighborhoods we serve, FAQ (3 area-specific questions), Call to action
    4. Uniqueness check: 3 points on what sets this page apart
    5. Missing information that would make the page stronger
    
    End by listing assumptions and any local rule I should confirm with the city before publishing.

AI search and citations

2 prompts

Audit how answer engines describe your category, build citation test sets and make key pages easier to quote accurately.

  1. #24AI SEOPerplexityWith example

    Get a hub-and-spoke internal link matrix pairing exact entity anchor text with required co-occurring concepts and flagging cannibalizing pages.

    Preview prompt
    Act as a technical SEO and entity-architecture specialist. I will paste my topic hub, a list of existing or planned spoke pages, and the core subject right after this prompt. If I have not pasted them yet, ask me for those three items in one numbered message and wait.
    
    Build a hub-and-spoke internal linking and entity plan for search and answer engines. Follow these rules:
    1. Assign one primary entity definition to each page so no two pages compete for the same concept.
    2. Write exact anchor text using the target entity name; never use generic text like learn more.
    3. Pair each link with 2 to 3 required co-occurring concepts that must appear in the surrounding paragraph.
    4. Flag any pages overlapping on primary entity intent and specify a fix: merge, redirect, or re-scope.
    5. Base every link on the pages I provide; do not invent URLs or traffic data.
    
    Output format:
    - Hub-and-Spoke Link Matrix: Markdown table with columns Source Page, Target Page, Exact Entity Anchor Text, and Required Co-Occurring Concepts.
    - Entity Cannibalization Flags: Bulleted list naming overlapping pages, the contested entity, and the fix.
    - Gaps and Assumptions: One short note on missing spoke topics or assumptions made from the input.
  2. #25SEOWith example

    AI Visibility Audit: Test Prompts, Mentions and Citations

    Get a 20-prompt test set and recording sheet for checking where AI answer engines mention or cite your brand, then a citation table, page fixes and an honest presence plan from your results.

    Preview prompt
    Role: You are an SEO analyst who specializes in AI visibility: how answer engines such as ChatGPT, Perplexity, Gemini and Google AI features mention and cite brands.
    
    Task: Phase 1: design a test prompt set and a recording sheet. Phase 2: once I paste results, analyze them and recommend actions.
    
    Start: Ask me these questions in one numbered message and wait for my answers:
    1. Brand, main products or services, and the URLs that should be cited
    2. 2 to 4 competitors
    3. Audience and topic, plus any real buyer questions you hear from customers
    4. Which tools and modes to test (for example ChatGPT with and without search, Perplexity, Gemini)
    5. Facts that must be stated correctly (pricing model, locations, key features)
    6. Country and language to test from, and runs per prompt (default 3)
    If I paste earlier results instead, skip Phase 1 and go to Phase 2. If I leave something out, list it under Missing information and continue.
    
    Constraints:
    - Do not claim to have run live queries or browsed. You cannot know current AI answers. Work only from results I paste and label everything else a hypothesis.
    - Answers vary by run, location and model version. Treat single runs as anecdotes.
    - Do not promise citations or ranking changes. No made-up sources or statistics.
    - Separate observed evidence, assumptions and recommendations.
    - Only recommend honest, disclosed contributions to third-party sites. Never suggest fake accounts, paid hidden reviews or planted mentions.
    
    Workflow:
    1. Phase 1: write 20 test prompts across five types: discovery, comparison, problem, brand and reputation. Add a recording sheet with columns for prompt ID, tool, mode, date, run, brand mentioned (Y/N), position, accuracy against the must-be-correct facts, cited URLs and competitors named.
    2. Phase 1: list the pages and entity facts to check before testing, as hypotheses.
    3. Phase 2: from the pasted results, build a citation table and count mention rate per prompt type only from that data. Show the counts.
    4. Phase 2: list inaccuracies and the most-cited domains, sorted into our pages, competitor pages and third-party pages.
    5. Phase 2: diagnose likely causes as hypotheses (content gaps, unclear entity data, few third-party mentions, blocked crawlers, thin comparison pages) and say how to verify each.
    
    Output format:
    1. Decision summary (max 6 lines)
    2. Missing information and assumptions
    3. Test prompt set (table: ID | Prompt | Type | Why it matters)
    4. Recording sheet (column headers and one blank example row)
    5. Phase 2 only: citation table (Prompt ID | Tool | Mentioned | Accuracy | Cited URLs | Source type) and findings table (Finding | Evidence rows | Confidence)
    6. Action list: Action | Page or entity | Hypothesis | Impact | Effort | How to verify on a re-test
    7. Phase 2 only: presence plan for third-party pages that keep being cited (Page | Who runs it | Honest contribution | Owner | Recheck date)
    8. Risks and next actions
    
    Quality checks:
    - All five prompt types are covered and read like real buyer questions.
    - Every finding cites a row of pasted data or is labeled a hypothesis.
    - Counts match the pasted data; accuracy is judged only against the facts I supplied.
    - Actions name specific pages or entities, not generic advice.

Playbook

How to get better results

Which AI model to pick

ChatGPT
Clustering Search Console exports, drafting content briefs and iterating on title and meta description variants.
Claude
Auditing long articles against competing pages, rewriting dense technical specs and keeping editorial briefs crisp.
Perplexity
Checking live search landscape citations, spotting which domains answer engines quote and verifying current SERP angles.

Follow-up prompts to refine any output

  1. Flag any keyword group above that risks cannibalising an existing page and recommend whether to merge or separate them.
  2. Turn the content brief into a strict writer checklist with exact questions to answer in the opening section.
  3. Convert the top technical findings into developer tickets with reproduction steps and verification criteria.

FAQ

About this list

Good SEO prompts never pretend an assistant has live search volume or ranking data. Instead, you paste exports from Search Console, a keyword tool or a site crawler, and the prompt clusters terms by searcher intent, flags cannibalisation, maps internal links and builds structured briefs tied to what already ranks.

You will also find technical triage prompts that turn noisy crawl sheets into prioritised engineering tickets, local SEO prompts for Google Business Profiles and reviews, plus modern AI search prompts for checking how often your brand is cited in answer engines. Each prompt includes an example output and opens straight in ChatGPT, Claude or Gemini.

Can ChatGPT provide accurate keyword search volume?

No. Language models cannot read live search volume or keyword difficulty scores from memory, and any numbers they invent will mislead your plan. Pull real queries and impressions from Google Search Console or a keyword tool first, paste that list into these prompts, and let the model handle intent grouping, topic clustering and brief writing.

Will AI-assisted SEO content get penalised by Google?

Google evaluates content on usefulness, originality and accuracy rather than whether AI helped draft an outline. Thin pages written without subject expertise struggle regardless of how they were produced. Use these prompts to research search intent, tighten structure, spot missing questions and refresh stale pages, while keeping first-hand examples, verified data and human editorial review in every article.

Which SEO tasks save the most time with ChatGPT?

Intent clustering, content brief creation, title and meta description testing, and turning crawl reports into prioritised engineering tasks usually save the most hours each week. Those jobs involve sorting large lists of URLs or queries into consistent templates, which chat models handle quickly once you supply clean exports and clear ranking criteria.

How do these prompts help with AI search and answer engine visibility?

Several prompts in this list focus on how answer engines discover and cite pages. They help you build a repeatable set of buyer questions to test across ChatGPT, Perplexity and Gemini, audit whether your pages answer core questions directly near the top, and add clear comparison tables, definitions and structured FAQ blocks worth quoting.

What data should I paste before running a technical SEO audit prompt?

Paste a sample export of affected URLs along with status codes, canonical targets, indexability flags and traffic tiers from your crawler or Search Console. You do not need to upload an entire domain at once; twenty to fifty representative rows let the prompt group root causes and draft clear acceptance criteria for developers.