35 production-tested Claude prompts covering AI Overview optimization, technical SEO, content strategy, keyword research, local SEO, link building, and analytics. The exact prompt library RankSages uses inside client retainers.

Claude has quietly become the most useful AI assistant for SEO work in 2026. Where ChatGPT excels at conversational tasks, Claude handles long contexts, structured outputs, and nuanced strategic thinking that match how real SEO work happens. The 35 prompts below are the exact ones we have tested across 240+ client engagements and now use inside RankSages to run our own client work across 240+ client engagements tracked between January 2025 and April 2026. They cover every major SEO function from AI Overview optimization through GSC anomaly investigation. Copy them, customize the bracketed variables, and you have a working SEO operations system.

Before you use any prompt: load business context first

Every prompt in this guide assumes Claude already knows your business. Spend 5 minutes loading context before running any prompt and your outputs will be 10x more useful. Paste this into Claude first:

You are my senior SEO consultant. Here is everything you need to know about my business so all future answers are tailored to my situation:

Company: [Your company name]
Industry: [Your industry]
Primary services: [List 3-5 main offerings]
Target customer: [Describe ideal customer in 2 sentences]
Geographic markets: [Cities or countries you serve]
Current monthly organic traffic: [GSC number]
Top 5 commercial keywords I rank for: [List with positions]
Top 3 competitors: [Domain names]
Current SEO tools: [Ahrefs, GSC, etc.]
Brand voice: [3 adjectives, e.g. confident, direct, no-fluff]
Single biggest goal for next 6 months: [Specific outcome]

Confirm you have this context, then wait for my first SEO task.
A quick note before you dive in:
I have been writing about SEO since 2010 and the version of this prompt library I shipped to clients in 2024 had 12 prompts. By Q1 2026 it grew to these 35 because AI Overview citation became its own discipline. Every prompt below has been used on at least 30 RankSages client engagements before it earned a spot here. If you find one that breaks for your industry, email me and I will fix it.

Table of Contents

  1. AI Search & Answer Engine Optimization (6 prompts)
  2. Technical SEO (5 prompts)
  3. Content Strategy (5 prompts)
  4. Keyword Research (5 prompts)
  5. Local SEO and Google Business Profile (5 prompts)
  6. Link Building and Digital PR (4 prompts)
  7. Analytics and Reporting (5 prompts)

By 2026, ranking in Google AI Overviews, ChatGPT search, and Perplexity citations is no longer a niche specialty, it is the baseline. These six prompts handle the core AI search optimization workflow.

Workflow diagram showing prompt chaining for SEO tasks
Chained prompts where each output feeds the next produce 10x the value of running prompts in isolation.

Technical SEO

Technical SEO is where AI assistants save the most time. These five prompts handle the tedious diagnostic work that used to require manual log file analysis and crawl reports.

1. AI Overview Citation Gap Audit

Identifies which AI Overviews your competitors are cited in that you are not, the single highest-ROI AI search opportunity.

Prompt
Act as a senior SEO consultant specializing in Google AI Overviews.

I will paste a list of 20 commercial keywords from my industry. For each keyword, I want you to:

1. Predict whether the keyword is likely to trigger an AI Overview (Y/N) based on query type
2. If yes, list the 3-5 information sub-questions the AI Overview would synthesize an answer from
3. For each sub-question, describe what kind of page content would most likely be cited (definition page, comparison table, case study, listicle, expert quote, statistic)
4. Identify which sub-questions my existing site content [LIST 5 EXISTING URLS] is best positioned to answer
5. Flag which sub-questions I have NO content for, these are my citation gaps
6. Output as a table: keyword | likely AI Overview | sub-questions | my current content match | content gaps

Keywords: [PASTE KEYWORDS HERE, ONE PER LINE]

Why this matters: AI Overview citation share is now the single biggest predictor of AI search traffic. This prompt finds the exact content gaps where you are losing citation opportunities to competitors and tells you what type of content would close each gap.

2. Passage Optimizer for AI Citation

Rewrites a single page section to maximize its chances of being extracted by AI Overview, ChatGPT, and Perplexity.

Prompt
Act as an AI search optimization specialist.

I will paste a section from one of my blog posts. Rewrite the opening passage of that section so it has the highest possible chance of being cited by Google AI Overviews, ChatGPT, and Perplexity.

Apply these specific rules:
- Open with a self-contained 50-80 word direct answer to the question implied by the heading
- The opening passage must make sense if pulled out of context and shown standalone
- Include exactly one specific data point, number, or named example
- Use active voice and present tense
- Avoid hedge words ("might", "could", "perhaps")
- Avoid em dashes
- Avoid generic closing phrases, weak hedge transitions, and time-bound digital-marketing opener clichés
- Match RankSages brand voice: confident, direct, no fluff

After the rewrite, explain:
1. What you changed and why
2. Which AI Overview citation factor each change addresses
3. Whether the section still flows naturally into the rest of the content

Section to rewrite:
[PASTE SECTION HERE]

Why this matters: AI Overview citation is decided at the passage level. A single optimized passage can earn citations across 5-30 related queries, multiplying your AI search visibility per hour of work invested.

3. AEO Content Brief Generator

Produces a complete content brief optimized for AI Overview citation, not traditional SERP ranking.

Prompt
Act as a senior content strategist specializing in AEO (Answer Engine Optimization).

Build a complete content brief for an article targeting the keyword: [KEYWORD]

The brief must include:

1. Primary target query and 8-12 secondary queries this article should also rank/cite for
2. Search intent classification (informational, commercial, transactional, navigational)
3. Article angle that differentiates from the top 5 ranking pages [LIST URLS]
4. Recommended word count and section count
5. H2 structure where each H2 is phrased as a question a real searcher would ask
6. For each H2, write a 60-word "anchor passage" that directly answers the question and is optimized for AI Overview extraction
7. One specific original data point, framework, or named example required per section
8. Internal linking targets (3-5 existing pages this article should link to)
9. Schema requirements (Article, FAQPage if applicable, BreadcrumbList, Person)
10. E-E-A-T signals to include (author credentials, expert review, citation sources)
11. Three meta title variants and three meta description variants

Output as a structured brief with clear section headings, ready to hand to a writer.

Why this matters: A traditional content brief optimizes for ranking. An AEO brief optimizes for AI Overview citation, which is a different game with different writing patterns. This brief format produces content that wins on both axes.

4. Knowledge Graph Entity Audit

Diagnoses how strong your brand entity signals are, the single most underrated AI search ranking factor.

Prompt
Act as a technical SEO specialist focused on entity SEO and Knowledge Graph optimization.

Audit my brand entity profile and identify gaps that limit my AI search visibility.

For my brand [BRAND NAME] (domain: [DOMAIN]), check:

1. Does my brand have a Wikidata entry? If yes, list its QID. If no, flag as critical gap.
2. Does my brand have a Wikipedia article? If no, what would be required to qualify (notability criteria)?
3. List the sameAs URLs I should have in my Organization schema (LinkedIn, Crunchbase, official social profiles, industry directories)
4. Identify which of my top 3 competitors [LIST COMPETITORS] have stronger entity signals and what they have that I do not
5. Check whether my founder/CEO has a personal Knowledge Graph presence (Wikidata QID, author profiles on major publications, speaking engagements indexed)
6. Recommend 5-10 specific actions, ordered by impact, to close my entity gaps within 90 days
7. For each action, estimate the difficulty (low/medium/high), time required, and expected entity strength impact

Output as a prioritized action plan with deadlines.

Why this matters: Entity grounding is the single biggest reason similar content from a stronger brand gets cited while yours does not. Closing entity gaps moves the needle on both AI Overview citation and traditional search rankings.

5. AI Search Query Forecasting

Predicts which queries will start triggering AI Overviews in the next 6 months so you can pre-optimize.

Prompt
Act as a Google AI Overview pattern analyst.

I will give you 30 keywords I currently rank for. Predict which ones are MOST likely to start triggering an AI Overview within the next 6 months that currently do not.

Use these signals to score each keyword:
- Query type: informational and comparative queries trigger AI Overviews first
- Query length: 4+ word queries are more likely than 1-2 word queries
- Commercial intent: pure transactional (e.g. "buy X near me") rarely triggers AI Overviews
- Existing SERP features: queries with Featured Snippets are pre-disposed
- Topic category: YMYL topics (health, finance, legal) and how-to queries lead adoption

For each of the 30 keywords:
1. Score AI Overview adoption likelihood: high / medium / low
2. Estimated timeframe (next 30 days / 90 days / 180 days / unlikely)
3. Specific content preparation I should do now to be cited when it adopts
4. Whether my current ranking position [I WILL PROVIDE] is strong enough to be in consideration

Output as a sortable table prioritized by combined likelihood and current ranking strength.

Keywords:
[PASTE 30 KEYWORDS WITH CURRENT POSITION]

Why this matters: AI Overview adoption is rolling out in waves. Optimizing a page for AI Overview citation 60 days before the AI Overview launches on that query makes you the default cited source on day one.

6. Schema Markup for AI Citation

Generates the full schema stack (Article + Person + Organization + FAQ + BreadcrumbList) for a single page.

Prompt
Act as a structured data specialist.

Generate complete JSON-LD schema markup for this page that maximizes its eligibility for AI Overview citation, ChatGPT browsing citations, and Google Featured Snippets.

Page details:
URL: [PAGE URL]
Title: [PAGE TITLE]
Author: [AUTHOR NAME, JOB TITLE]
Author bio: [2 SENTENCES]
Author LinkedIn: [URL]
Published: [DATE]
Last updated: [DATE]
Word count: [NUMBER]
Topic: [ONE SENTENCE]
Includes FAQ section: [Yes/No, if yes paste questions and answers]
Article category: [CATEGORY]

Output requirements:
1. Article schema with full E-E-A-T signals (author with knowsAbout array, sameAs links, jobTitle)
2. Person schema for the author with all available signals
3. Organization schema for the publisher with sameAs links
4. FAQPage schema if FAQ section exists, with all questions and answers
5. BreadcrumbList schema reflecting the actual URL structure
6. speakable property highlighting the most citation-worthy passages
7. All schema combined into a single JSON-LD block, validated for Google Rich Results

Validate the output mentally against Google's Article structured data requirements and Person schema best practices.

Why this matters: Pages with 4+ schema types are cited in AI Overviews at 2.3x the rate of pages with 1-2 schema types. This prompt produces production-ready schema in one pass instead of three.

Content Strategy

Content strategy in 2026 is no longer about volume, it is about building topical authority that AI search engines recognize. These prompts handle the strategic content decisions.

7. Crawl Budget Diagnosis from Sitemap + GSC

Identifies pages Google is wasting crawl budget on so you can free up capacity for important content.

Prompt
Act as a technical SEO consultant specializing in large-site crawl budget optimization.

I will provide:
- My XML sitemap URL: [SITEMAP URL]
- My GSC Coverage report showing pages crawled but not indexed: [PASTE LIST]
- My GSC Crawl Stats showing the URLs Googlebot hits most often: [PASTE TOP 50]

Analyze and report:

1. Categorize each crawled-not-indexed URL: thin content / duplicate / wrong canonical / accidental noindex / parameter URL / archive page / feed / other
2. Identify the top 5 URL patterns wasting the most crawl budget
3. Recommend specific fixes: which to noindex, which to canonicalize, which to 410, which to merge
4. Estimate crawl budget freed up by each fix
5. Prioritize fixes by impact: critical (do this week) / important (do this month) / cleanup (do later)
6. Generate the exact robots.txt rules, X-Robots-Tag headers, and canonical tag changes needed
7. Suggest 5 high-priority new pages that should be added to the sitemap because they are missing

Output as a structured action plan with implementation snippets.

Why this matters: “Crawled but not indexed” is the most common GSC issue in 2026 and is almost always a quality and crawl budget problem, not a technical bug. This prompt finds the exact pages dragging down your site quality score.

8. INP Optimization Plan

Diagnoses why your INP (Interaction to Next Paint) is failing and produces a prioritized fix list.

Prompt
Act as a Core Web Vitals specialist focused on INP (Interaction to Next Paint, the metric that replaced FID in March 2024).

My site's current INP at 75th percentile: [VALUE in ms]
Target: under 200ms
URL of worst-performing page: [URL]

Analyze the likely root causes based on common INP failures in 2026:
1. Heavy event handlers (click, input, keydown) running synchronously
2. Long tasks blocking the main thread during user interactions
3. Large React/Vue/Angular component re-renders triggered by clicks
4. Third-party scripts (chat widgets, ad pixels, analytics) hijacking the main thread
5. CSS containment not configured on interactive elements
6. Excessive DOM size making layout calculations slow
7. Missing requestIdleCallback or scheduler.yield for non-urgent work

For each likely cause:
- Explain the technical mechanism
- Suggest the diagnostic to confirm (Chrome DevTools steps)
- Provide the specific code fix or configuration change
- Estimate INP improvement potential (small/medium/large)

Produce a prioritized 30-day INP optimization plan ordered by impact-per-hour-of-effort.

Why this matters: Most sites still use 2023 Core Web Vitals advice that references FID, which was retired. INP failures are different problems requiring different fixes, especially on heavy JavaScript sites.

9. JavaScript SEO Audit

Diagnoses whether your JavaScript-rendered content is actually visible to Googlebot and AI crawlers.

Prompt
Act as a JavaScript SEO specialist.

My site is built on [Next.js / Nuxt / React / WordPress with heavy JS / Other], rendered as [SSR / SSG / CSR / Hybrid].

Page to audit: [URL]

Run a JavaScript SEO audit answering these questions:

1. What content is in the initial HTML response (visible to Googlebot crawler without rendering)?
2. What content is injected via JavaScript only?
3. Are critical SEO elements (title, meta description, canonical, hreflang, robots meta, JSON-LD schema) in the raw HTML or JS-injected?
4. If JS-injected, what is the risk that Googlebot will not see them?
5. Are there any JavaScript-injected canonical tags that conflict with raw HTML canonical tags?
6. Does the page return HTTP 200 status code? (Google will not render JS on non-200 pages)
7. Are AI crawlers (GPTBot, ClaudeBot, PerplexityBot) able to see the content? They generally do not execute JavaScript.

For each issue found:
- Explain the SEO impact
- Provide the specific fix (SSR vs prerender vs hybrid approach)
- Indicate priority level

Reference the December 2025 Google JavaScript SEO documentation update.

Why this matters: AI crawlers in 2026 mostly do not render JavaScript. A site that ranks fine on Google but has zero AI search visibility almost always has critical content trapped behind JS rendering.

10. Robots.txt and Sitemap Strategy

Produces an optimal robots.txt and sitemap configuration for your specific CMS and crawl priorities.

Prompt
Act as a technical SEO architect.

Build me an optimal robots.txt and sitemap strategy for:

Site: [DOMAIN]
CMS: [WordPress / Shopify / Webflow / Custom]
Total pages: [APPROXIMATE]
Page types I have: [services / blog posts / location pages / products / etc]
Pages I do NOT want indexed: [list]
AI crawler policy: [allow all / allow Google + Bing only / block AI training crawlers]

Produce:

1. Complete robots.txt file with proper user-agent blocks for: Googlebot, Bingbot, GPTBot, ChatGPT-User, ClaudeBot, PerplexityBot, Google-Extended, Bytespider, CCBot
2. Specific disallow rules to block low-value URLs (parameter URLs, search results, login pages, admin paths, comment feeds)
3. Sitemap structure recommendation: should I use a sitemap index? How should I split sitemaps by content type?
4. Specific sitemap exclusions (which pages should NOT be in the sitemap even if indexable)
5. Crawl-delay recommendations if any
6. IndexNow setup instructions for Bing/Yandex/Naver fast indexing
7. Validation steps to confirm the configuration is working

Output the robots.txt as a ready-to-deploy file and the rest as an action checklist.

Why this matters: A misconfigured robots.txt is one of the few single-file mistakes that can tank an entire domain in search. A well-configured one frees crawl budget for the pages that earn revenue.

11. Site Architecture Restructure Plan

Diagnoses information architecture problems and proposes a new URL structure that improves crawl depth and topical clustering.

Prompt
Act as an information architect specializing in SEO-driven site structure.

Audit my site's current architecture:
Domain: [DOMAIN]
Total pages: [NUMBER]
Current URL patterns:
[LIST EXAMPLES OF EACH PAGE TYPE WITH URL PATTERN]

Current navigation structure:
[PASTE MAIN NAV HIERARCHY]

Run an architecture audit and propose improvements:

1. Identify pages currently more than 3 clicks deep from the homepage (crawl depth issue)
2. Identify orphan pages (no internal links pointing to them)
3. Identify topic clusters that are scattered across non-related URL paths
4. Identify URL patterns that signal weak hierarchy to Google
5. Identify navigation patterns that hide important content from crawlers

Propose a restructured architecture:
- New URL structure with hub-and-spoke clusters
- Specific 301 redirect map from old URLs to new URLs
- Internal linking changes needed (which pages should link to which)
- Navigation menu changes
- Sitemap restructure
- Expected impact: crawl depth improvement, topical authority concentration, ranking lift estimate

Output as a phased migration plan with risk mitigation steps.

Why this matters: Most sites accumulate architectural debt over years. A single architecture refresh can lift 30-60% of pages by 2-5 ranking positions because Google understands the topical structure better.

Keyword Research

Keyword research has shifted from finding high-volume terms to finding queries with conversion intent and AI Overview citation potential. These five prompts do exactly that.

12. Topic Cluster Architect

Builds a complete hub-and-spoke content cluster around a strategic topic, optimized for both Google and AI search citation.

Prompt
Act as a senior content strategist specializing in topical authority.

I want to dominate the topic of [BROAD TOPIC, e.g. "AI search optimization"] over the next 6 months.

Build a complete topic cluster:

1. Identify the single best pillar page topic (broadest, most search volume, most evergreen)
2. List 8-12 cluster page topics that support the pillar
3. For each cluster page:
   - Recommended title
   - Primary keyword + 5 secondary keywords
   - Estimated search volume (rough range)
   - Search intent (informational / commercial / etc)
   - Word count target
   - 3-5 H2 section ideas
   - Which other cluster pages should it internally link to
4. Internal linking map: pillar -> clusters, cluster -> cluster
5. Publishing sequence over 12 weeks (which to publish first, second, etc.)
6. AI Overview citation potential for each piece
7. Identify content gaps where this cluster differs from competitors' coverage of the same topic

Output as a structured plan with a visual hierarchy of pillar and clusters, plus a 12-week publishing calendar.

Why this matters: Google’s 2026 ranking model heavily favors sites with deep topical coverage. A well-designed cluster can rank 50+ pages simultaneously while individual posts on the same topics rarely break the first page.

13. Content Brief Builder for Long-Form Articles

Produces a 2,000-word article brief specific enough for any writer to execute without strategic questions.

Prompt
Act as a senior content editor producing a complete brief for a freelance writer.

Article assignment: a 2,500-word piece on [TOPIC]

Build a brief that includes:

1. One-sentence article angle that differentiates from competitors
2. Target audience (job title, seniority, what they already know)
3. Primary keyword + 8-15 secondary keywords with rough volumes
4. Search intent and content format (guide / listicle / comparison / case study / framework)
5. Required sections (H2s) with 2-3 sentence summary of what each section must cover
6. Three competitor URLs to study + one-line note on what each does well that we should beat
7. Specific data points or examples the writer must include (with source URLs)
8. Internal links required (3-5 existing URLs on my site, with anchor text suggestions)
9. External authoritative sources to cite (3-5 with URLs)
10. Tone: [adjectives]
11. Avoid: [specific words, phrases, AI-generation tells]
12. Three working title options
13. Three meta description options
14. CTA placement and copy

Output as a structured brief ready to send to a writer with no follow-up questions needed.

Why this matters: A vague brief produces a vague article. A specific brief produces an article that ranks. This prompt forces the brief to be specific enough that the writer has no strategic decisions left to make.

14. Content Refresh Prioritization

Tells you which existing posts to update first for the biggest traffic lift with the least effort.

Prompt
Act as a content optimization strategist.

I will provide a list of my existing blog posts with these metrics for each:
- URL
- Title
- Current monthly traffic from GSC
- Current average position
- Top query that drives traffic
- Date published
- Date last updated

Build a content refresh priority list:

1. Score each post on refresh value (1-10) using:
   - Traffic potential if it ranked one position higher
   - Outdated information likelihood (older posts on fast-moving topics)
   - Current ranking position (positions 4-15 have highest lift potential)
   - Commercial intent (transactional posts > informational for revenue lift)
   - Backlinks (don't damage posts that already earn links)

2. For the top 10 priority posts:
   - Estimated traffic gain after refresh
   - Specific refresh recommendations (what content to add/remove/update)
   - Whether to maintain the URL or 301 to a new URL
   - Internal linking changes needed
   - Schema updates needed
   - Whether to update the dateModified or write a "2026 update" note

3. Output as a ranked refresh queue with estimated ROI per post

Posts:
[PASTE POSTS WITH METRICS]

Why this matters: Most agencies refresh content based on what feels old. This prompt refreshes based on what will return the most traffic per hour invested, often 5-10x better ROI than gut-driven refreshes.

15. Pillar Page Outline Generator

Produces a complete 4,000-word pillar page outline designed to rank, get cited in AI Overviews, and link out to your cluster pages.

Prompt
Act as a pillar page specialist.

Build a complete outline for a 4,000-word pillar page targeting: [PRIMARY KEYWORD]

The pillar must:
- Rank for the primary keyword and 15-25 related queries
- Be the authoritative reference on the topic for AI Overview citation
- Internally link to my existing cluster pages: [LIST 6-10 URLS]
- Match RankSages brand voice: confident, direct, data-backed

Produce:

1. Working title (with the primary keyword)
2. Subtitle / hook paragraph (under 50 words)
3. Table of contents with 10-15 H2 sections
4. For each H2:
   - Phrased as a question a real searcher asks
   - 60-word anchor passage optimized for AI Overview extraction
   - Bullet list of supporting points (3-7 items)
   - Which cluster page to internally link to from this section
   - Whether the section needs a chart, table, or diagram
5. Final FAQ section with 8-12 Q&As
6. CTA placement strategy (top, middle, bottom)
7. Required schema: Article + Person + Organization + FAQPage + BreadcrumbList
8. Meta title + meta description options
9. Three featured image concept ideas
10. Internal linking checklist: which cluster pages must link back to this pillar

Output as a structured outline ready to hand to a writer.

Why this matters: A great pillar page is the single biggest-impact piece of content on a site. It can drive 10,000+ visits per month while serving as the authority hub that lifts every related cluster page.

16. Content Decay Recovery Plan

Diagnoses why a once-high-traffic post is losing traffic and produces a recovery plan.

Prompt
Act as a content recovery specialist.

This post is losing traffic month over month:
URL: [URL]
Current monthly traffic: [NUMBER]
Peak monthly traffic (when): [NUMBER, DATE]
Decay pattern: [steady decline / cliff drop / seasonal / unclear]
Top query that used to drive traffic: [QUERY]
Current ranking position for that query: [POSITION]
Date last meaningfully updated: [DATE]

Diagnose the decay:

1. Identify the most likely root cause:
   - Algorithm update (which one)
   - Competitor created better content
   - Content went stale (outdated stats, broken examples, removed product features)
   - Search intent shifted
   - SERP feature ate the clicks (AI Overview, Featured Snippet, People Also Ask)
   - Keyword cannibalization from new content on same site
   - Site-wide quality signal decline
   - Lost backlinks

2. For each likely cause, suggest the diagnostic to confirm

3. Produce a recovery plan with specific actions, expected timeline, and probability of success

4. If recovery is not realistic, recommend whether to: leave it / merge into another post / 410 it / redirect it

Reference recent Google algorithm updates in your analysis.

Why this matters: Most sites accumulate decaying content that drags down crawl budget and overall site quality. Recovering or removing it lifts the entire domain, often boosting unrelated pages by 5-10%.

Local SEO and Google Business Profile

Local SEO in 2026 still drives most service-business revenue. These prompts handle Google Business Profile optimization and local ranking factors.

17. Commercial Intent Keyword Mining

Finds the queries that actually convert, not just the ones that get traffic.

Prompt
Act as a commercial keyword researcher.

My business: [DESCRIBE IN 2 SENTENCES]
Primary services: [LIST]
Average customer value: [NUMBER]
Target geographic markets: [LIST]

Brainstorm 50 commercial-intent keywords I should rank for, organized into:

1. Bottom-of-funnel buying-intent (e.g. "best [service] in [city]", "[product] vs [competitor]")
2. Solution-aware comparison (e.g. "[my service] vs [alternative approach]")
3. Problem-aware research (e.g. "how to [solve problem my service solves]")
4. Branded comparison (e.g. "my brand vs competitor brand")
5. Pricing intent (e.g. "[service] cost", "how much does [service] charge")

For each keyword:
- Estimated search volume range (high / medium / low)
- Estimated keyword difficulty (1-10)
- Search intent specificity (clear / moderate / vague)
- Commercial value to my business (1-10)
- Quick win potential (1-10) based on ranking difficulty vs commercial value
- Recommended content type to rank for it

Sort by quick win potential and output as a prioritized table.

Why this matters: Most keyword research lists are full of high-volume informational queries that bring traffic but no revenue. This prompt finds the queries with actual buying intent that drive pipeline.

18. Long-Tail Question Discovery

Surfaces the exact questions your audience asks that have low competition and high commercial signal.

Prompt
Act as a long-tail keyword researcher specializing in question-format queries (which are the queries most likely to trigger AI Overviews).

Topic: [BROAD TOPIC]
Audience: [TARGET CUSTOMER]

Generate 40 long-tail question keywords organized by:

1. "How do I" questions (operational)
2. "What is the best way to" questions (decision)
3. "Why does X happen" questions (diagnostic)
4. "When should I" questions (timing)
5. "How much does X cost" questions (pricing)
6. "Is X worth it" questions (validation)
7. "X vs Y" questions (comparison)
8. "Can I" questions (feasibility)

For each question:
- Underlying search intent
- Whether it likely triggers an AI Overview
- Estimated difficulty to rank
- Whether my business is positioned to answer credibly
- Suggested content format

Flag the 10 highest-priority questions that combine low competition, high commercial relevance, and high AI Overview citation potential.

Why this matters: Question keywords convert 2-4x better than statement keywords because the user has already identified a specific problem. They also trigger AI Overviews at the highest rate of any query format.

19. Competitor Keyword Gap Analysis

Identifies the exact keywords competitors rank for that you do not, ordered by commercial value.

Prompt
Act as a competitive keyword analyst.

My domain: [DOMAIN]
Top 3 competitors: [LIST DOMAINS]
Industry: [INDUSTRY]

I will provide:
- A list of keywords each competitor ranks in positions 1-20 for
- My current rankings for the same keywords (or zero if I don't rank)

Analyze the gap:

1. Identify keywords where ALL three competitors rank in top 20 but I don't rank at all (highest priority gaps)
2. Identify keywords where 2 of 3 competitors rank in top 20 but I don't (second priority)
3. For each gap keyword, classify by:
   - Search volume tier
   - Commercial intent
   - Content format ranking on page 1 (article / product / video / comparison)
   - Estimated difficulty
   - Strategic value to my business

4. Rank the gaps by ROI (combination of volume, commercial intent, achievability)

5. For the top 15 gap keywords:
   - Suggest article angle that beats the competitors
   - Recommend whether to write new content or update existing content
   - Identify which competitor we should specifically target to displace

Output as a prioritized table with the 15 highest-ROI gaps highlighted.

Data:
[PASTE COMPETITOR KEYWORD DATA]

Why this matters: Competitor gap analysis is the single highest-confidence keyword opportunity source because the data proves the keyword has commercial value and is rankable for sites in your category.

20. Featured Snippet and PAA Opportunity Scout

Finds queries where you almost own the Featured Snippet or People Also Ask and need one optimization to capture it.

Prompt
Act as a SERP feature specialist.

For each of my pages [LIST PAGES with current top queries], identify:

1. Queries where I rank in positions 2-5 and a Featured Snippet exists, owned by a competitor
   - Specific snippet format (paragraph / list / table / video)
   - What the competitor's winning passage looks like
   - The specific edit my page needs to take the snippet

2. Queries where I rank in positions 1-10 and a People Also Ask box exists
   - Which PAA questions I could realistically answer
   - The 40-80 word direct answer to add to my page for each question

3. Queries where I rank in positions 1-10 and a Knowledge Panel exists
   - Whether my entity could legitimately be added to the Knowledge Panel
   - The Wikidata/Wikipedia steps required

Output as a prioritized opportunity list with the exact text changes needed to each page.

Why this matters: Featured Snippet takeover is one of the few SEO tactics that produces a 100-300% traffic increase from a single 30-minute edit. Most opportunities are sitting one paragraph rewrite away from happening.

21. Semantic Keyword Clustering

Groups a flat keyword list into semantic clusters, each one a candidate for a single piece of content.

Prompt
Act as a semantic keyword clustering specialist.

I will paste a flat list of 100-300 keywords from my industry. Group them into semantic clusters where each cluster represents queries that should be targeted by a single piece of content (not multiple pieces, which would cause cannibalization).

For each cluster:
1. Cluster name (the most representative query)
2. All keywords in the cluster
3. Combined estimated search volume
4. Average difficulty
5. Recommended single URL/content piece to target the cluster
6. Search intent classification
7. AI Overview trigger likelihood

Rules:
- A query should appear in exactly one cluster
- Cluster only when the SERPs would significantly overlap (same top 5 results for both queries)
- Flag any cluster where ranking for one query would conflict with ranking for another (intent mismatch)
- Identify clusters where multiple existing pages on my site already compete (cannibalization risk)

Keywords:
[PASTE LIST]

Why this matters: Most keyword tools dump raw lists that look impressive but cause cannibalization when targeted directly. This prompt produces a clean architecture where every keyword has a clear home.

Link building is the discipline most improved by AI assistance because it is fundamentally a communication task. These four prompts handle the outreach and ideation workload.

22. GBP Category Optimization

Picks the optimal primary and secondary categories for your Google Business Profile, the single biggest local ranking lever.

Prompt
Act as a Google Business Profile optimization specialist.

My business:
Name: [BUSINESS NAME]
Industry: [DESCRIBE WHAT YOU DO]
Services offered: [LIST]
Service area: [CITIES OR RADIUS]
Customer types: [B2B / B2C / both]
Top 5 services by revenue: [LIST]
Current GBP primary category: [CATEGORY]
Current GBP secondary categories: [LIST]

Recommend:

1. The optimal PRIMARY category (this drives 60-70% of local ranking signals, choose carefully)
2. 5-9 secondary categories that maximize coverage without diluting the primary signal
3. For each recommendation, explain why this category matches the services that drive revenue
4. Flag any current categories that should be removed (they dilute relevance)
5. Identify gaps where Google does not have a perfect category for one of my services, and how to handle (which closest category to choose)
6. Local competitor analysis: what categories are my top 3 competitors using and what does that signal about market norms

Format as a recommended GBP category configuration with implementation steps.

Why this matters: Primary GBP category is the single biggest local ranking factor and the most under-optimized. Switching to the right primary category can lift map pack rankings by 3-7 positions in 30 days.

23. Local Citation Audit

Identifies NAP inconsistencies and missing citation sources that weaken your local authority.

Prompt
Act as a local citation specialist.

My business NAP:
Name: [EXACT NAME]
Address: [EXACT ADDRESS]
Phone: [PHONE]
Website: [DOMAIN]
Industry: [INDUSTRY]
Service area: [CITIES]

Run a citation audit:

1. List the top 25 citation sources my business should be on, prioritized by authority and relevance to my industry/location
2. For each citation source, indicate whether I should:
   - Manually claim and optimize
   - Use a service like Yext or BrightLocal
   - Skip (low value for my industry)
3. Identify the 10 most common NAP variations that cause citation inconsistency:
   - Phone number formats (with vs without country code, hyphens, parentheses)
   - Address variations (Street vs St, Suite vs Ste, ZIP+4 vs ZIP)
   - Business name variations (LLC vs Limited, Inc vs Incorporated)
4. Recommend the single canonical NAP format I should use everywhere
5. Suggest 5 industry-specific citation sources beyond the standard top 25
6. Estimate the impact on local pack rankings from completing this audit

Output as an action checklist with priority order.

Why this matters: NAP consistency across 25-50 authoritative citations is still a meaningful local ranking factor in 2026. Sites with strong citation profiles rank in the local pack at 2-3x the rate of sites with weak citation profiles.

24. Review Response Templates by Sentiment

Produces response templates for positive, neutral, and negative reviews that strengthen local SEO and customer trust.

Prompt
Act as a customer experience strategist who understands the SEO impact of review responses.

My business: [DESCRIBE BUSINESS]
Brand voice: [3 ADJECTIVES]
Common positive themes in our reviews: [LIST]
Common complaints in negative reviews: [LIST]

Write response templates for these scenarios:

1. 5-star review with no detail ("Great service!") - 3 variations
2. 5-star review with specific praise about a service/employee - 3 variations
3. 4-star review with mild constructive feedback - 3 variations
4. 3-star review citing one specific issue - 3 variations
5. 2-star or 1-star review with valid complaint - 3 variations
6. 1-star review that seems unfair or possibly fake - 3 variations

For each template:
- Use natural language (not corporate)
- Include keyword opportunities where natural (city, service name) without being spammy
- Acknowledge the specific feedback when present
- Avoid generic phrases like "we value your feedback"
- Include a next-step CTA when appropriate
- Keep responses under 100 words

Also provide 5 best practices for review response timing and authority signaling.

Why this matters: Review responses are read by Google AND prospects. Responses optimized for both create stronger local rankings AND higher conversion rates from review readers.

25. Local Landing Page Brief

Produces a complete brief for a city-specific landing page that ranks without triggering thin-content penalties.

Prompt
Act as a local SEO content strategist.

Build a brief for a landing page targeting: [SERVICE] in [CITY]

The page must:
- Rank for "[service] [city]" and 10-15 related local queries
- Avoid being thin or duplicate of my other location pages
- Provide genuine local value to the visitor
- Match RankSages brand voice

Produce:

1. Recommended URL slug
2. H1 title
3. Hero section: 2-3 sentence intro that mentions the city specifically and why this service matters in this market
4. 6-8 H2 sections, each with:
   - Heading
   - 80-150 word section summary
   - Local data, examples, or context required (population, market trends, regulations, neighborhoods)
   - Internal linking opportunities

5. Required unique elements (no other location page should have these):
   - 1 local case study (with city-specific details)
   - 1 local data table or chart
   - 1 reference to city-specific regulations, neighborhoods, or industry conditions
   - 3-5 city-specific FAQ items

6. Schema requirements
7. Meta title and meta description
8. Internal linking targets: which national service pages and which other location pages this page should link to
9. CTAs and conversion elements

Output a brief that produces a page genuinely different from any other location page on my site.

Why this matters: Thin templated location pages trigger Google’s Site Reputation Abuse and Helpful Content updates. This prompt produces location pages with enough unique content to rank legitimately.

26. Service Area Expansion Plan

Identifies which new cities to expand into based on opportunity score, not gut feel.

Prompt
Act as a local market expansion strategist.

My business:
Current service area: [CITIES]
Service type: [B2B / B2C / both]
Average customer value: [NUMBER]
Customer acquisition cost: [NUMBER]
Top performing existing market: [CITY] - and why it works

Evaluate these 10 candidate expansion markets: [LIST 10 CITIES]

For each candidate market, score on:
1. Market size for my service (population in target demographic)
2. Local search demand for my service (high / medium / low)
3. Local competition strength (number of established players, average authority)
4. Local SEO difficulty (citation density, review competitiveness)
5. Cultural / regulatory fit with my business model
6. Logistical feasibility (can I actually deliver service there)
7. Estimated time to first conversion
8. Estimated 12-month revenue potential

Produce:
- Total opportunity score per market (1-100)
- Ranked list with the top 3 markets and why they win
- For each top 3 market, the specific 90-day SEO launch plan
- Markets to deprioritize and why

Output as a strategic expansion roadmap.

Why this matters: Random geographic expansion is the most expensive growth strategy in service businesses. Data-driven expansion produces 3-5x better ROI per market entered.

Analytics and Reporting

The analytics work most agencies do badly is interpretation. Charts are easy. Knowing what to do based on the chart is hard. These five prompts handle interpretation.

27. HARO / Qwoted Pitch Generator

Writes a journalist-quality pitch in under 5 minutes that has a 30-50% acceptance rate.

Prompt
Act as a senior PR professional pitching a journalist for a HARO or Qwoted query.

The query:
Publication: [PUBLICATION NAME]
Journalist: [NAME if available]
Query topic: [PASTE FULL QUERY]
Deadline: [DATE]

My credentials and angle:
Expertise: [YOUR EXPERTISE IN 2 SENTENCES]
Specific experience relevant to this query: [DETAILS]
Original data or insight I can provide: [DETAILS]
Quotable angle (counterintuitive, specific, fresh): [ANGLE]

Write the pitch:

1. Subject line under 60 characters that signals "I read your query and have a specific answer"
2. Opening that proves you read the query (specific reference, not generic)
3. The core insight or quote in 80-150 words, written in journalist-ready voice (no marketing fluff, no jargon)
4. 1-2 supporting data points or examples
5. Brief credentials proof (1-2 lines)
6. Easy-to-use callback options (phone, time zones, willingness for follow-up)
7. Total pitch length under 200 words

Avoid:
- Marketing speak
- Empty marketing adjectives (the over-used innovation/industry-leadership cliché family)
- Generic opening like "I saw your query"
- Bulleted credentials lists
- Long signature blocks

Output the pitch in journalist-ready format.

Why this matters: HARO and Qwoted pitches that look like marketing emails are deleted in seconds. Pitches that look like a useful response from a smart human get quoted. The difference is mostly writing style, which AI handles well when prompted correctly.

28. Broken Link Building Outreach

Drafts the outreach email for a broken link building campaign with high enough specificity to get a response.

Prompt
Act as a link building specialist conducting broken link outreach.

Context:
Target page (the page with the broken link): [URL]
Target site authority: [APPROX DA/DR]
Target audience: [WHAT TYPE OF SITE - blog, association, university, etc]
Broken link URL: [BROKEN URL]
Original page topic: [WHAT IT WAS ABOUT]
My replacement content: [URL TO MY CONTENT]
Why my content is a legitimate replacement: [2 SENTENCES]

Write an outreach email:

1. Subject line that signals you are helpful, not asking for a favor
2. First sentence that proves you read their page specifically
3. The broken link issue described in 1-2 sentences with the specific anchor text or section where it appears
4. Your replacement suggestion offered as helpful information, not as a request
5. Why your content actually fits (acknowledge if it does not perfectly fit)
6. No-pressure close that acknowledges they may not be the right person or may have other priorities
7. Email length under 150 words
8. Friendly, peer-to-peer tone

Also provide 2 follow-up email templates if no response after 7 days and after 14 days.

Avoid:
- Generic "I found a broken link" openings
- Begging or excessive politeness
- Long pitches about your content
- Tracking pixels
- Multiple links in the email

Why this matters: Most broken link outreach reads like cold sales pitches and gets ignored. Outreach that reads like a helpful note from a peer gets 12-20% response rates and 5-8% link placement rates.

29. Resource Page Outreach

Crafts the email for getting added to a resource page or “useful links” list.

Prompt
Act as a digital PR specialist.

Target:
Resource page URL: [URL]
Resource page topic: [TOPIC]
Site owner (if known): [NAME]
Type of resources currently listed: [TYPE]
What makes my resource genuinely fit this list: [REASON]

My resource:
URL: [MY URL]
Description in 1 sentence: [DESCRIPTION]
What it provides that the existing resources do not: [UNIQUE VALUE]
Free / paid: [STATUS]

Write the outreach email:

1. Subject line that signals you are offering value, not asking
2. Opening that proves you have actually used or studied their resource page
3. Brief context about why you are reaching out specifically to them
4. Your resource offered with 1-2 specific reasons it fits their list (NOT generic praise of your own content)
5. Acknowledge any reason it might not fit and let them decide
6. No-pressure ask: would you consider adding it
7. Sign-off with one-line credential
8. Email length under 175 words

Tone: peer-to-peer, helpful, never desperate.

Avoid:
- Mass-personalization tells (no merge field artifacts)
- "I love your content" opener
- Excessive praise
- Asking for a "quick favor"

Why this matters: Resource pages are still one of the cleanest link sources in 2026 because they have actual editorial standards. The right outreach gets featured permanently on pages with 10+ year link equity.

30. Linkable Asset Ideation

Generates 15-20 specific linkable asset ideas that journalists, bloggers, and curators would actually link to.

Prompt
Act as a digital PR strategist specializing in linkable asset development.

My business: [DESCRIBE]
Industry: [INDUSTRY]
Audiences who might link to me: [JOURNALISTS, BLOGGERS, ACADEMICS, ASSOCIATIONS, ETC]
Existing data I have access to: [INTERNAL DATA, CLIENT DATA, ETC]
Existing assets I have: [REPORTS, TOOLS, RESEARCH]

Generate 15-20 linkable asset ideas. For each asset:

1. Asset title (specific, intriguing, citable)
2. Asset format (original research, free tool, ultimate guide, dataset, interactive calculator, industry survey, definitive list)
3. Who would link to it (specific publications, blogs, communities)
4. Production complexity (low / medium / high)
5. Estimated time to produce
6. Expected backlink potential (small / medium / large)
7. Single sentence on the angle that makes it linkable (data, controversy, utility, novelty)

Rank the 15-20 ideas by ROI (backlink potential divided by production complexity).

Avoid ideas that are:
- Generic "ultimate guide" formats with no unique data
- Round-up posts that depend on contributor goodwill
- Tools that already exist in better form elsewhere
- Research that does not produce a quotable headline statistic

Why this matters: A single great linkable asset can earn 100-500 backlinks over its lifetime. Most “content marketing” budget is wasted on assets that no one would ever link to. This prompt forces brutal honesty about linkability.

31. GSC Anomaly Investigation

Diagnoses sudden traffic changes in Google Search Console with the same logic a senior consultant would use.

Prompt
Act as a senior SEO analyst investigating a GSC anomaly.

The anomaly:
Date range with abnormal pattern: [DATES]
What changed: [traffic up / traffic down / impressions changed / position changed / specific page]
Magnitude of change: [PERCENT]
Affected page or section: [URL OR PATTERN]

I will paste relevant GSC data: [PASTE DATA SCREENSHOT DESCRIPTION OR EXPORT]

Run the diagnosis:

1. Classify the anomaly type (algorithm update, technical issue, content change, competitor change, seasonal, search behavior shift)
2. For each possible cause, list the diagnostic check that would confirm or rule it out
3. Cross-reference Google's confirmed algorithm update calendar for the date range
4. Check for technical issues (indexation drop, crawl errors, ranking volatility for affected pages)
5. Investigate competitor activity (new content, large updates, schema changes)
6. Examine the affected pages for: recent content changes, technical changes (template changes, CMS updates), backlink changes
7. Output a likely cause ranked list (top 3 candidates with probability)
8. Recommend specific next-step diagnostic actions for the top 2 candidates
9. If the cause is confirmed, recommend the recovery action
10. Set expectations for recovery timeline

Be specific about which diagnostic tool to use for each check (GSC, Ahrefs, Wayback Machine, etc).

Why this matters: Most agencies guess at the cause of GSC anomalies and end up wasting hours on irrelevant fixes. Systematic diagnosis identifies the real cause in 30-60 minutes.

32. Monthly SEO Client Report

Produces a client report that actually explains what happened, what it means, and what you are doing about it.

Prompt
Act as a senior SEO consultant writing a monthly report for a non-technical client.

Client business: [DESCRIBE]
Reporting period: [MONTH]
Top 3 client priorities: [LIST]

I will provide the data:
- GSC: total clicks, impressions, position, top queries, top pages, gainers, losers
- GA4: organic sessions, conversions, conversion rate
- Backlinks: new referring domains
- Rankings: top 20 tracked keywords with month-over-month change

Produce the report:

1. Executive summary (2-3 paragraphs, written for a CEO, no jargon, focus on revenue impact and progress against goals)
2. Key wins this month (3-5 specific outcomes with the work that produced them)
3. Key challenges (be honest, do not hide losses, explain what happened)
4. Metric overview table (clicks, impressions, position, conversions with month-over-month %)
5. Top 5 ranking gains (keyword, old position, new position, what we did)
6. Top 5 ranking declines (keyword, old position, new position, what we are doing about it)
7. Content shipped this month (titles, URLs, early performance)
8. Technical fixes shipped (specific issues fixed and impact)
9. Next month focus (3-5 priorities with rationale)
10. Strategic question for the client (something they need to decide that affects work next month)

Tone: confident, transparent, peer-to-peer (not deferential). Length: 2 pages of single-spaced text.

Why this matters: Agency reports that hide losses or pad with vanity metrics break client trust. Reports that show the real story (wins, losses, what we are doing about it) build the trust that retains clients for years.

Related deep-dive: Running a one-person operation? Our small business SEO guide shows how to compress this entire workflow into a 3-hour weekly cadence.

33. Traffic Forecasting Model

Builds a 12-month organic traffic forecast based on current trajectory and planned work.

Prompt
Act as an SEO forecasting analyst.

Current state:
Current monthly organic clicks: [NUMBER]
Current monthly impressions: [NUMBER]
Current average position: [NUMBER]
Last 12 months trend: [growing / declining / flat with %]
Top ranking pages and their current positions: [PASTE TOP 20]

Planned work for next 12 months:
- New content pieces: [NUMBER]
- Existing pages to refresh: [NUMBER]
- Technical improvements: [LIST MAJOR ONES]
- Link building target: [NEW REFERRING DOMAINS PER MONTH]

Build a 12-month forecast:

1. Monthly traffic projection broken into:
   - Existing pages continuing to perform (with attrition factor)
   - Existing pages improved by refreshes (with lift estimates)
   - New pages launching and ramping up
2. Confidence intervals (best case / base case / worst case)
3. Key assumptions and their sensitivity (which assumption, if wrong, breaks the forecast)
4. Risk factors that could derail the forecast (algorithm updates, competitor activity, seasonality)
5. Leading indicators to track monthly that would tell us if the forecast is on track
6. Recommended forecasting cadence (weekly check / monthly check / quarterly recalibration)

Output as a forecast model with explicit math.

Why this matters: Most traffic forecasts are wishful thinking. A forecast with explicit assumptions and confidence intervals creates accountability and lets you actually know whether work is on track.

34. Conversion Path Analysis

Investigates why your traffic does or does not convert, separating SEO problems from CRO problems.

Prompt
Act as a conversion analyst combining SEO and CRO expertise.

My business: [DESCRIBE]
Primary conversion: [DEFINE]
Current monthly organic traffic: [NUMBER]
Current monthly organic conversions: [NUMBER]
Conversion rate from organic: [%]
Industry benchmark conversion rate: [% IF KNOWN]

I will provide:
- Top 10 landing pages by organic traffic, each with: traffic volume, conversion rate, average time on page, bounce rate
- Top 10 conversion paths (which pages users visit before converting)
- Top exit pages

Investigate:

1. Identify pages with high traffic but low conversion (traffic problem? intent mismatch? page UX problem? offer problem?)
2. Identify pages with low traffic but high conversion (where to send more traffic, how to amplify)
3. Diagnose intent mismatches (where SEO is bringing wrong-fit visitors)
4. Diagnose conversion friction (which pages need CRO work, what specifically)
5. Identify high-converting pages that could be replicated for other topics
6. Recommend 5-10 specific actions, separated into "SEO actions" (change traffic) and "CRO actions" (change conversion)
7. Estimate revenue impact per action

Output as a prioritized roadmap with clear ownership (SEO team vs design/dev team).

Why this matters: Most organic traffic problems are diagnosed as SEO problems when they are actually CRO problems, or vice versa. This prompt separates them so each team works on the right issues.

35. SEO Health Score Calculator

Produces a single composite health score for your domain across 8 dimensions, plus the actions that move the score.

Prompt
Act as a senior SEO auditor producing a comprehensive SEO health assessment.

Domain: [DOMAIN]

I will provide data for each dimension. Score each dimension on a 1-10 scale and produce an overall weighted SEO health score.

Dimensions:

1. Technical SEO (weight 18%): crawlability, indexability, mobile, Core Web Vitals, security
2. Content quality (weight 20%): depth, E-E-A-T signals, uniqueness, freshness
3. On-page SEO (weight 12%): titles, meta, headings, internal linking, schema
4. Off-page authority (weight 15%): referring domains, anchor profile, brand mentions
5. AI search readiness (weight 12%): AI Overview citation share, entity grounding, AI crawler accessibility
6. Local SEO (weight 8% or 0% if not local): GBP, citations, reviews, local content
7. User experience (weight 10%): bounce rate, time on site, conversion path quality
8. Reporting and measurement (weight 5%): tracking accuracy, analytics setup, GSC health

For each dimension:
- Score 1-10 with rationale
- Top 3 issues found
- Recommended actions ordered by impact

Produce:
- Overall weighted SEO health score (0-100)
- Single biggest-impact action (the action that moves the score most)
- 30-day improvement plan
- 90-day improvement plan
- Score targets to reach in 90 days

Why this matters: A composite health score makes SEO progress legible to leadership in a way single-metric reports never can. It also forces honest assessment across dimensions teams normally ignore.

RankSages tools that pair with these prompts

These prompts produce strategy. The tools we built below produce the data those strategies need. Used together they form a complete operating system.

Free Tool

AI Visibility Checker

Pulls real Ahrefs ranking data and identifies which of your queries trigger Google AI Overviews. Pairs with Prompts 1, 2, and 5.

Free Tool

Schema Markup Generator

Generates Article + Person + Organization + FAQPage JSON-LD in one pass. Pairs with Prompt 6.

Free Tool

Keyword Research Tool

Real-time search volume and difficulty data. Feeds Prompts 17 through 21.

Free Tool

SEO Audit Tool

Site-wide technical audit output you can feed into Prompts 7 through 11.

How to actually use these 35 prompts

Reading 35 prompts is interesting. Running 35 prompts is transformative. Here is the suggested rollout if you want to integrate them into your workflow:

  1. Week 1: Load the business context block once. Save it as a reusable snippet. This single step is the biggest force multiplier in the entire system.
  2. Week 1: Run prompts 31 (GSC anomaly), 35 (SEO health score). These tell you where you actually stand right now.
  3. Week 2: Run prompts 17, 19, 20 (keyword research and gap analysis). Build your priority opportunity list.
  4. Week 3: Run prompts 12 and 15 (cluster architecture, pillar page). Build the content strategy.
  5. Week 4: Run prompts 1, 4, 6 (AI Overview audit, entity audit, schema). Lay the AI search foundation.
  6. Week 5 onward: Use prompts 13, 14, 16 weekly to manage content production and refresh cycles. Use prompts 27-30 monthly for link building. Use prompt 32 monthly for client reporting.

What this prompt library does not replace

Strategic judgment, client relationships, original research, technical implementation, and the hundreds of small decisions that make SEO work. These prompts are accelerators, not replacements. The best SEO consultants in 2026 use AI assistants for the speed and depth but keep the strategy human.

If you want to skip the DIY route entirely and have RankSages run this system on your domain, we apply these exact workflows inside our retainer engagements. Every plan includes AI Search Optimization at no additional cost.

All claims and benchmarks in this guide draw from three datasets RankSages maintains internally:

  • RankSages AI Overview Tracker: 240+ client domain audits, 8,400+ tracked keywords, weekly AI Overview citation captures since January 2025.
  • RankSages Prompt Benchmark Q1 2026: 47 SEO prompts tested across 4 leading AI assistants (Claude, GPT-5, Gemini Pro, Grok), scored on output usefulness and strategic depth.
  • RankSages Client Outcomes Database: Aggregated SEO performance data from 240+ retainer engagements, January 2025 through April 2026.

All percentage figures cited as “in our data” or “across our portfolio” reference these three datasets. Cited correlations are Pearson coefficients calculated against binary citation outcomes (cited or not cited per query).

Prompt engineering principles that apply to all 35 prompts

Each of the 35 prompts above uses the same underlying prompt engineering principles. Understanding these principles lets you adapt the prompts to your specific situation, write your own prompts for use cases not covered here, and debug prompts that produce weak outputs.

Principle 1: Role assignment before task

Every prompt opens with a role assignment (“Act as a senior SEO consultant”, “Act as a Google AI Overview analyst”). This is not a formality. Claude’s output quality varies measurably based on the role context. A role assignment changes which patterns of expertise the model draws on, which vocabulary it uses, and how confident or hedged its conclusions are.

Best practices: be specific about seniority (junior vs senior produces meaningfully different outputs), specialty area (technical SEO vs content strategy vs analytics), and context (in-house vs agency consultant).

Principle 2: Output structure requirements explicit upfront

Saying “produce a table” produces a table. Saying “produce a structured brief with headings” produces a brief. Saying nothing produces unpredictable formatting. Every prompt above specifies output structure (table, bulleted list, numbered steps, code blocks) because unstructured outputs are harder to consume downstream.

Best practices: name the exact output format you want. Mention any sections that must be included. Specify length range when relevant (“under 200 words”, “8-12 items”). Be explicit about format conventions (“use markdown headers”, “format as JSON”).

Principle 3: Constraint specification (what NOT to do)

Constraints are often more useful than positive instructions. “Avoid generic SEO advice” produces better outputs than “Give specific SEO advice” because Claude understands what to filter out. Many prompts above include explicit avoid lists.

Best practices: tell Claude what categories of output to avoid. Reference common failure modes. Forbid specific phrasings if you have brand voice rules.

Principle 4: Example inputs and variables in brackets

Bracketed variables ([KEYWORD], [DOMAIN], [INDUSTRY]) signal what the user will customize. They also make the prompt self-documenting. Users immediately see which inputs they need to provide.

Best practices: use square brackets consistently. Make variable names descriptive. Include format hints where useful ([DATE in YYYY-MM-DD format]).

Principle 5: Expertise grounding through specific frameworks

Prompts that reference specific frameworks (“AEO/GEO”, “INP not FID”, “Pearson correlation”) produce more expert outputs than generic prompts. The framework references signal to Claude that the expected output should use the same level of specificity.

Best practices: name the frameworks, methodologies, or terminology you want the output to use. Reference current state of the art (2026 standards, not 2022). Be explicit about the audience level (beginner vs expert).

How to write your own prompts for use cases not covered here

The 35 prompts above cover the most common SEO workflow steps. Inevitably you will encounter use cases that require new prompts. The template for writing reliable new prompts:

Step 1: Identify the input and output

What information will you give Claude? What output do you need back? Be specific about both. Vague inputs produce vague outputs.

Step 2: Write the role assignment

What expertise should Claude bring to this task? Senior SEO consultant? Content editor? Technical SEO specialist? Match the role to the task.

Step 3: Specify the task in 1-3 sentences

What exactly should Claude do with the input? Be precise. “Analyze this content” is vague. “Identify the 5 most likely reasons this content fails to rank in top 10, ordered by probability” is precise.

Step 4: Define output structure

What format should the output take? Table? Numbered list? Bulleted analysis? Structured brief? Code snippet? Specify clearly.

Step 5: Add constraints and avoid lists

What should the output NOT include? Generic advice? Hedge language? Specific phrasings that do not match your brand voice? Be explicit.

Step 6: Test with 3 different inputs

Run the prompt with three different inputs to verify it produces consistent output quality across variations. If outputs vary wildly in usefulness, the prompt needs more specificity.

Step 7: Document and version

Save the working prompt with a clear name and use-case description. As you refine it, keep version history so you can compare outputs over time.

The 12 prompt templates we use most outside of SEO

The same prompt engineering principles work for non-SEO tasks too. The templates that produce consistently useful outputs in our internal use across operations, sales, and client communication:

  1. Client weekly status update: structured report covering wins, challenges, blockers, and next steps. Used by every account manager weekly.
  2. Sales call follow-up email: pulls the 3-5 most important points from a call transcript into a structured follow-up that proves you listened.
  3. Proposal scope document: turns a discovery call into a structured scope of work with deliverables, timelines, and pricing.
  4. Client objection response: drafts thoughtful responses to common client objections that acknowledge concerns and provide context.
  5. Internal documentation: turns informal Slack threads or meeting notes into clean documentation pages.
  6. Job description writer: drafts role descriptions that capture actual job requirements without buzzword inflation.
  7. Interview question generator: produces interview questions calibrated to specific role requirements and seniority levels.
  8. Performance review framework: structured review templates that focus on observable behaviors not personality traits.
  9. Decision documentation: captures the context, options considered, decision made, and reasoning for major team decisions.
  10. Customer support template library: response templates for common support scenarios that maintain consistent tone.
  11. Investor update structure: monthly or quarterly update template covering metrics, narrative, asks, and team news.
  12. Board meeting preparation: pre-read documents that present material at the right level of abstraction for board members.

Each of these follows the same engineering principles as the SEO prompts: role assignment, structured output, explicit constraints, bracketed variables, expertise grounding. Once you internalize the pattern, you can write effective prompts for any structured business task.

When to NOT use AI prompts (the human-only zones)

Honest disclosure: there are work categories where AI prompts hurt more than help. These are the human-only zones we maintain in our agency:

Crisis communication

Client penalty recovery, major algorithm response, communication after a security incident. These require human judgment, authentic empathy, and accountability that AI-assisted communication compromises.

Sales objection handling in real-time

Live client conversations require pattern recognition AI does not replicate well. Pre-thinking objection responses (using the templates above) is fine; reading AI-generated responses in real-time is not.

Hiring decisions

AI can draft interview questions and analyze candidate written responses. The hiring decision itself should be human because it involves judgment about cultural fit, growth potential, and long-term contribution that requires human assessment.

Strategic pivots

Decisions to enter or exit a market, change pricing, restructure team. These need human strategic thinking informed by tacit knowledge AI does not have access to.

Conflict resolution

Inter-team conflicts, client relationship problems, partner disputes. Resolution requires authentic human engagement that AI mediation cannot substitute for.

The teams that get the most leverage from AI prompts are the ones that draw these lines clearly. They use AI aggressively for the productive zones and protect the human-only zones from creeping AI substitution.

FAQ

Are these prompts only for Claude or do they work with other AI assistants?

In our testing, they work with any modern AI assistant including GPT-5, Gemini Pro, and Grok. We tested all 35 prompts across the top four assistants and Claude produced the most strategically sound outputs in our internal benchmarks of 47 prompts tested across 4 leading AI assistants in Q1 2026, particularly for long-form content briefs and competitive analysis. For pure data tasks like keyword clustering, the differences narrow significantly.

How long does it take to get useful output from these prompts?

Most prompts produce useful output in 30-90 seconds. The longest (cluster architecture, pillar page outline) can take 2-3 minutes. The biggest time saver is loading business context once at the start of a session so subsequent prompts are automatically tailored to your business.

Should I trust the data outputs (volume estimates, ranking predictions, etc)?

Treat AI-generated estimates as directional, not precise. Always verify volume estimates with your keyword research tool (Ahrefs, Semrush) and ranking data with Search Console. The strategic frameworks are reliable, the specific numbers should be validated against real data sources.

Can I customize these prompts for my specific industry?

Yes, and you should. Each prompt has bracketed variables for customization. For deeper customization, add industry-specific context to the initial business context block. For example, healthcare SEO has compliance requirements (HIPAA, AHPRA) that should be added so all outputs respect them automatically.

Will using AI for SEO trigger Google penalties?

No. Google penalizes content that lacks helpfulness, originality, and expertise. Whether that content was written with AI assistance is irrelevant. AI-assisted strategy, briefs, and drafts that are edited by humans with real expertise are indistinguishable from fully human-written content in search performance.

How often should I re-run these prompts?

Recommended cadence: GSC anomaly investigation weekly, content briefs and link building outreach as needed for active projects, health score quarterly, cluster strategy refresh every 6 months, competitor gap analysis monthly.

Related deep-dive — SEO ROI Calculator: After running these prompts, model the revenue lift with our free ROI calculator. Read more →

Related deep-dive — 2026 SEO Glossary: Quick reference for every AEO, GEO, and AI Search term these prompts use. Read more →

Frequently asked questions

Are these prompts only for Claude or do they work with other AI assistants?

In our testing, they work with any modern AI assistant including GPT-5, Gemini Pro, and Grok. We tested all 35 prompts across the top four assistants and Claude produced the most strategically sound outputs in our internal benchmarks of 47 prompts tested across 4 leading AI assistants in Q1 2026, particularly for long-form content briefs and competitive analysis. For pure data tasks like keyword clustering, the differences narrow significantly.

How long does it take to get useful output from these prompts?

Most prompts produce useful output in 30-90 seconds. The longest (cluster architecture, pillar page outline) can take 2-3 minutes. The biggest time saver is loading business context once at the start of a session so subsequent prompts are automatically tailored to your business.

Should I trust the data outputs (volume estimates, ranking predictions, etc)?

Treat AI-generated estimates as directional, not precise. Always verify volume estimates with your keyword research tool (Ahrefs, Semrush) and ranking data with Search Console. The strategic frameworks are reliable, the specific numbers should be validated against real data sources.

Can I customize these prompts for my specific industry?

Yes, and you should. Each prompt has bracketed variables for customization. For deeper customization, add industry-specific context to the initial business context block. For example, healthcare SEO has compliance requirements (HIPAA, AHPRA) that should be added so all outputs respect them automatically.

Will using AI for SEO trigger Google penalties?

No. Google penalizes content that lacks helpfulness, originality, and expertise. Whether that content was written with AI assistance is irrelevant. AI-assisted strategy, briefs, and drafts that are edited by humans with real expertise are indistinguishable from fully human-written content in search performance.

How often should I re-run these prompts?

Recommended cadence: GSC anomaly investigation weekly, content briefs and link building outreach as needed for active projects, health score quarterly, cluster strategy refresh every 6 months, competitor gap analysis monthly.