Claude Prompts for SEO: 25 Copy-Paste Prompts Built Properly

You are a senior SEO analyst. I am pasting my Google Search Console query export below. Analyse the demand I already have instead of inventing keywords.

<gsc_data>
[paste your Queries.csv rows here]
</gsc_data>

Task: group these queries into topic clusters, and for each cluster report: total impressions, total clicks, impression-weighted average position, and the single highest-leverage query.

Constraints: use only the numbers in the data. Do not estimate search volumes or difficulty; if I need those, tell me to validate in a keyword tool. Do not create a cluster with fewer than 2 queries.

Format: one table (Cluster | Impressions | Clicks | Weighted position | Top query), then a 3-bullet read of the biggest opportunity.

Verify: before answering, check the maths on one cluster and show the check. If the paste looks truncated, ask me for the rest instead of proceeding.

You are an SEO strategist mapping queries to pages. Here are my target queries:

<queries>
[paste 10-30 queries]
</queries>

Task: classify each query's intent (informational, commercial, transactional, navigational) and assign the page type that wins it (guide, comparison, product, category, tool, FAQ).

Constraints: if a query is ambiguous, say so and give both readings rather than forcing one. No generic advice paragraphs.

Format: table (Query | Intent | Page type | One-line rationale). End with any queries that should share one page instead of getting separate pages.

Verify: flag every query where you are less than confident, so I can check the live results myself.

You are an AI search strategist. I want the map of questions an AI engine explores when someone asks about my topic.

<topic>
[your topic or target query]
</topic>

<audience>
[who searches this and what they are trying to do]
</audience>

Task: fan this topic out into the related questions an AI engine would pull answers for: comparisons, how-tos, costs, risks, alternatives, "for [situation]" variants. Group them by the single page that should own each group.

Constraints: questions must be ones a real person would ask, in their words. No invented search volumes. If a group is too thin to deserve its own page, fold it into another and say so.

Format: table (Question group | The questions | Page that should own it), then the 3 groups to build first.

Verify: mark any group you are unsure real searchers ask, so I can validate it before building.

You are a content researcher building a coverage matrix. I am deciding what my page must include to beat the current results for "[target query]".

<competitor_content>
[paste the headings or full text of the top 2-3 ranking pages, labelled A, B, C]
</competitor_content>

Task: build the coverage matrix: every distinct angle, claim, or subtopic those pages cover, classified as consensus (all cover it) or partial (some do). Then add the angles a subject expert would expect that none of them touch.

Constraints: classifications come only from the pasted content, no guessing what a page "probably" says. The missing-angles list is where I will differentiate, so make it specific to this query, not generic "add more examples" filler.

Format: table (Angle | A | B | C | Status), then the missing angles ranked by how much a searcher would care.

Verify: name any angle where you inferred coverage rather than saw it in the paste.

You are a demand researcher. Below is raw language from my actual customers.

<customer_language>
[paste sales emails, support tickets, call notes, or reviews, anonymised]
</customer_language>

<seed>
[your core topic or service]
</seed>

Task: extract every question and problem phrasing that relates to my seed topic, deduplicate, and group by funnel stage: still researching, comparing options, ready to act.

Constraints: keep the customers' own wording as the canonical phrasing; do not translate it into marketing language. Flag anything that still identifies a person. No invented questions.

Format: three funnel-stage lists in the customers' words, with a closing call per group: FAQ answer, dedicated page, or a section on an existing page.

Verify: flag any question that appears only once, so I know it is a single data point rather than a pattern.

You are a content strategist who refuses to write consensus content. I want to outrank the current results for "[target query]".

<competitor_headings>
[paste the H2/H3 headings from the top 3 ranking pages]
</competitor_headings>

<what_i_know>
[2-5 bullets: your real experience, data, or contrarian view on this topic]
</what_i_know>

Task: build an outline that covers what ranking pages agree on in less space, and leads with what none of them have, drawn from my inputs.

Constraints: every H2 phrased as a question. The unique material goes in the first third, never the end. If my "what I know" bullets are too thin to create real difference, say so plainly.

Format: H1 + H2/H3 skeleton, with one line under each H2 stating what the section proves.

Verify: mark each section [CONSENSUS] or [NEW] so I can see the gain at a glance.

You are a content strategist writing a brief another writer could execute without asking questions.

<target>
Query: [target query]
Searcher intent: [what the searcher is trying to do]
Audience: [who they are]
</target>

<my_angle>
[2-5 bullets: the experience, data, or position that makes this page different]
</my_angle>

Task: produce a complete brief: working title, the angle in one sentence, H2/H3 skeleton with one line per section on what it must prove, entities and terms the page must cover, internal links to include, and the evidence the writer needs to gather before starting.

Constraints: the unique angle leads the brief and the outline's first third. If my angle bullets cannot carry a differentiated page, stop and say what would. State the word range the job needs, no padding for its own sake.

Format: one brief document with labelled sections, ending with a pre-flight checklist for the writer.

Verify: read the brief back as the writer and list any question they would still have to ask, then answer it in the brief.

You are diagnosing a page whose growth stalled. Here is the page and its search data:

<page>
[paste the page copy]
</page>

<gsc_queries>
[paste the page's queries: query, clicks, impressions, position, ideally this period vs last]
</gsc_queries>

Task: diagnose why the page stopped growing and prescribe the refresh: sections to add, rewrite, or cut, plus any title or heading changes, each tied to a query in the data.

Constraints: every prescription must trace to evidence in the paste. If the data says the page is fine and the topic simply peaked, say "leave it alone" and stop; do not prescribe a rewrite to justify the exercise.

Format: diagnosis in 3 bullets, then a prioritised change list (Change | Evidence | Effort), then the one change to make first.

Verify: list the queries you could not explain with the current content, so I can decide whether they deserve a new page instead.

You are building a comparison page a reader could fact-check. Here is the real material:

<option_a>
[name + real specs, pricing, limits, your hands-on notes]
</option_a>

<option_b>
[the same for the second option]
</option_b>

<context>
[your relationship to the products: affiliate, vendor, neutral, and who the reader is]
</context>

Task: structure the comparison: the criteria that actually decide the choice, the comparison table, the honest verdict logic (who should pick A, who should pick B), and where my relationship to the products must be disclosed.

Constraints: use only the pasted specs and mark gaps [NEEDS DATA] instead of filling them. The verdict must follow from the table, so no crowning a winner the criteria do not support. If the options are near-identical, say a comparison page is the wrong format.

Format: page skeleton with the table, the verdict section, and the disclosure placement.

Verify: list every claim in the skeleton a reader could challenge, with the spec that backs it.

You are a ruthless line editor. Below is a draft that reads machine-written.

<draft>
[paste the draft]
</draft>

Task: edit it so a practitioner would believe a human wrote it, keeping every factual claim intact.

Constraints: delete every sentence that says nothing. Replace hedges ("can potentially help") with plain claims or cut them. Kill formulaic transitions ("In today's digital landscape", "It's important to note"). Do not add new facts. Keep my numbers exactly as written.

Format: the edited draft, then a short list of the patterns you removed so I stop writing them.

Verify: reread your edit and cut 10% more.

You are restructuring content to be quotable. Here is a section and the query it should answer:

<section>
[paste the section]
</section>

<query>
[the question this section targets]
</query>

Task: rewrite the section so the first sentence answers the query directly, in the query's own grammar, with the key term in bold. Evidence, nuance, and caveats follow the answer instead of preceding it.

Constraints: keep every fact; move them, do not cut them. The opening answer must stand alone if quoted with nothing around it. No throat-clearing ("When it comes to...").

Format: the rewritten section, then the standalone answer sentence on its own line so I can see exactly what an AI engine would lift.

Verify: read the first sentence alone. If a searcher who saw only that would still be unsure, rewrite it and show both versions.

You are an editor adding credibility a reader can verify. Here is my draft and my actual experience:

<draft>
[paste the draft]
</draft>

<credentials>
[real, checkable material: years doing this, projects or clients, data you gathered, mistakes you made, tools you use daily]
</credentials>

Task: weave the experience into the draft where it does argumentative work: first-person observations attached to the claims they support, real examples replacing hypothetical ones, and the limits of my experience stated plainly.

Constraints: use only what is in my credentials; invented experience is worse than none. Do not bolt on an "about the author" paragraph, the signals belong inside the argument. Flag every claim that still stands on nothing as [NEEDS PROOF].

Format: the revised draft with the additions in bold, then the [NEEDS PROOF] list.

Verify: confirm every added first-person statement traces to my credentials paste.

You are rewriting SERP packaging. My page ranks for "[query]" at position [X] with a low click-through rate.

<current>
Title: [current title]
Meta: [current meta description]
</current>

Task: write 5 title options and 3 meta options that carry the query's exact words near the front.

Constraints: titles under 60 characters, metas under 155. No clickbait the page cannot keep. No pipes-and-brand-only titles.

Format: numbered options, each with a one-line "why this earns the click".

Verify: for each title, state the character count. Flag any option where the promise exceeds what my page delivers.

You are writing SERP packaging in batch. Here are my pages:

<pages>
[one per line: URL | title | primary query | what the page offers]
</pages>

Task: write one meta description per page that carries the primary query's words early and gives a concrete reason to click.

Constraints: under 155 characters each. No "Learn more about...", no "Welcome to", no promise the page cannot keep. Each meta must differ meaningfully from the others, not one template with swapped nouns.

Format: table (URL | Meta | Character count).

Verify: recount the characters on the three longest and show the counts. If a page's input was too thin to write an honest meta, flag it instead of padding it.

You are auditing how machines read my page's structure. Here is the page:

<page>
[paste the page copy with its headings marked, or just the heading list]
</page>

<query>
[the primary query the page targets]
</query>

Task: audit the heading hierarchy: one H1 that carries the query, H2s phrased as the questions searchers ask, no skipped levels, and sections ordered by importance to the searcher rather than by writing order. Then propose the corrected skeleton.

Constraints: keep headings honest to the content beneath them; never rename a section to promise something it does not deliver. If a section answers no plausible question, recommend cutting it rather than reheading it.

Format: issues list (Issue | Heading | Fix), then the full revised H1-H3 skeleton.

Verify: under each revised H2, state the search query it now answers. An H2 with no query gets flagged, not defended.

You are a structured-data specialist. Here is my page content:

<page>
[paste the visible page copy]
</page>

Task: write a 4-6 question FAQ answering what a reader of this page still asks, then produce the matching FAQPage JSON-LD.

Constraints: every answer must be supported by the page content or the inputs I gave you; mark anything unsupported as [NEEDS SOURCE] rather than inventing it. Schema text must mirror the visible answers exactly.

Format: the FAQ as headings and paragraphs, then one JSON-LD block.

Verify: confirm the JSON parses and every schema answer matches its visible twin.

You are a structured-data specialist who marks up only what exists. Here is my product page:

<product_page>
[paste the visible product page copy: name, price, availability, description, review content if genuinely on the page]
</product_page>

Task: produce Product JSON-LD with offers, built from the visible values.

Constraints: every value must appear on the page. Missing values become [NEEDS VALUE], never a guess. No aggregateRating unless real review data is in the paste; invented ratings are the fastest route to a manual action.

Format: one JSON-LD block, then the [NEEDS VALUE] gaps and where each value should come from.

Verify: confirm the JSON parses, then walk each field back to the sentence in my paste it came from.

You are a technical SEO lead triaging a crawl. Here is my crawl export:

<crawl>
[paste rows: URL, status, title, meta, canonical, indexability]
</crawl>

Task: group every issue by type, then rank the groups by likely traffic impact, not by count.

Constraints: use only what is in the data. A 404 on a page nobody links to is not priority one. Say which issues need a human decision versus a mechanical fix.

Format: prioritised list, each with: issue, affected URLs (count + 3 examples), impact reasoning, the fix.

Verify: state which columns were missing from my paste that would change your ranking, so I can re-export properly.

You are a crawler, reading my robots.txt exactly as written. Here it is:

<robots>
[paste your robots.txt]
</robots>

<concerns>
[optional: what you are worried about, e.g. "is anything blocking AI crawlers?"]
</concerns>

Task: explain what each rule actually does, then flag the gaps between intention and effect: rules that block more than intended, AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended) allowed or blocked without a stated decision, and directives that do nothing.

Constraints: read the file literally, as a crawler would, without assuming what I meant. Distinguish "blocked from crawling" from "blocked from indexing", because robots.txt only controls the first.

Format: rule-by-rule table (Rule | What it does | Risk), then a corrected file if changes are needed.

Verify: state which user-agents you were unsure how to match, rather than guessing their behaviour.

You are building a redirect map for a migration. Here are the URL sets:

<old_urls>
[paste the current URLs, one per line, ideally with titles]
</old_urls>

<new_urls>
[paste the new site's URLs, one per line, ideally with titles]
</new_urls>

Task: map every old URL to the new URL that serves the same intent. Where no equivalent exists, assign the best parent category or recommend a 410, and say which.

Constraints: match by intent, never by URL similarity alone. No chains (old to old to new), and no mapping everything to the homepage, which throws away the equity the map exists to keep.

Format: CSV-ready table (Old URL | New URL | Type: exact / intent / parent / 410 | Note), then the unmapped list ranked by traffic risk.

Verify: count both lists and confirm every old URL appears in the output exactly once.

You are hunting quick wins. Here are my queries ranking positions 4-15:

<striking_distance>
[paste query, page, impressions, position rows]
</striking_distance>

Task: pick the 5 highest-leverage rows and prescribe the packaging fix for each: the title change, the heading to add, or the paragraph to answer the query directly.

Constraints: leverage = impressions x closeness to page one, not raw position. Prescriptions must be specific enough to paste, not "improve the content".

Format: for each of the 5: the query, the page, the diagnosis in one line, then the paste-ready fix.

Verify: if two queries on the same page compete, say which one the page should commit to and why.

You are an internal-linking strategist. Here are my site's pages and the new page I just published:

<pages>
[paste URLs + titles, one per line]
</pages>

<new_page>
[URL + 2-line summary of the new page]
</new_page>

Task: recommend 3-5 existing pages that should link TO the new page, with the exact sentence and anchor text for each placement, plus 2-3 pages the new page should link out to.

Constraints: anchors match the target page's title language, not "click here". Only suggest links a reader would genuinely follow. No footer or sidebar suggestions.

Format: table (From page | Suggested sentence with anchor | Why).

Verify: check no suggested anchor competes with the target page's own primary query on the source page.

You are a digital PR strategist who only pitches real stories. Here is what I have:

<assets>
[your data, survey results, unusual observations from your work, or a contrarian position you can defend]
</assets>

Task: generate 5 story angles a journalist would open, each anchored to my real material.

Constraints: no fake surveys, no invented statistics, no "study reveals" framing unless the study exists. If my assets are too thin for a story, say so and tell me what data would make one.

Format: for each angle: the headline a journalist would write, the hook sentence, the outlet type it fits, and which asset it uses.

Verify: mark any angle that needs additional data collection before it can be pitched honestly.

You are writing a Google Business Profile from verified facts. Here they are:

<business_facts>
[services, service area, years operating, real differentiators, themes from genuine reviews]
</business_facts>

Task: write the business description within Google's 750-character limit, front-loading what matters because only the opening shows before the "More" link. Then list the services with a one-line description each, and 5 Q&A pairs seeded from questions customers genuinely ask.

Constraints: every claim traces to my facts. No superlatives I did not earn ("best", "#1") and no keyword-stuffed sentences a human would wince at; the description is read by the person deciding whether to call.

Format: the description with its character count, the services list, then the 5 Q&As.

Verify: list anything in my facts you left out, and why.

You are deciding whether a local landing page deserves to exist, then outlining it. Here is the situation:

<service_location>
[the service + the location, e.g. "emergency plumbing in Footscray"]
</service_location>

<local_proof>
[real material for THIS location: jobs completed, local reviews, local staff, area-specific knowledge, pricing differences]
</local_proof>

Task: first judge whether my local proof can carry a page genuinely different from my other location pages. If yes, outline it: locally specific H2s, where each proof point goes, and the questions residents of this area actually ask. If no, say so and list the proof to gather first.

Constraints: nothing in the outline may be reusable for another suburb by swapping the place name; that is the doorway-page pattern that gets sites penalised. No invented local details.

Format: verdict first (build / do not build yet), then the outline or the proof-gathering list.

Verify: for each H2, state which proof point makes it specific to this location.