AI Search / GEO

The discipline of getting named and cited by AI answer engines. It is mostly a brand-positioning problem, not a technical one: roughly 85% of what AI says about you comes from off-site. You win by measuring per-model visibility, formatting content to be extractable, engineering consistent off-site signals, and running it as a continuous loop with attribution that can actually see AI-referred traffic.

When to use this

• Buyers in the category ask AI tools for recommendations and you want to be named, especially

for high-intent "best for X", comparison, and alternatives prompts.

• You rank on Google but never get pulled into AI answers.
• AI cites your content while recommending a competitor (ghost citations).
• You need a baseline of how ChatGPT / Perplexity / Gemini / Claude describe you today.
• Your analytics show AI doing "nothing" because AI-referred visits look like direct traffic.
• You need a sustainable monthly/biweekly GEO process, not a one-off.

When NOT to use this (reach for instead)

• For general content quality, E-E-A-T, and classic on-page SEO that feeds these citations:

use seo-content. Classic ranking still matters here (a #1 Google result is ChatGPT-cited ~43% of the time), so do not skip it.

• For writing the actual answer text, comparison copy, and positioning language: use

copywriting-messaging. This skill tells you the structure and placement; the words are a copy job.

• For building the GA4 attribution model, funnels, and reading the numbers: use

analytics-data. This skill gives the custom-channel regex and what to expect; the measurement plumbing is analytics.

How this works (decision path)

Follow the four-phase method: Diagnose -> Position -> Build -> Scale. Then maintain it.

1. Diagnose (measure first). You cannot fix what you have not measured, and models disagree

sharply (only ~11% domain overlap between ChatGPT and Perplexity), so test every major model. Run the query x model audit, log brand appearance, direct links, and ghost citations, then sort every gap into one of three types: coverage (rivals have multi-source reinforcement you lack), consistency (your messaging contradicts itself across sources), or specificity (you are described broadly while a rival owns a niche).

1. Fix by gap type.

• Ghost citations (cited but not recommended) -> a positioning/off-site fix, not more content.
• Consistency gap -> entity work: make one identical story everywhere.
• Answerability gap -> on-page: re-chunk content into extractable passages.
• Coverage/ecosystem gap -> earn off-site mentions and listicle placements.

1. Position + Build. Engineer the signals models read (narrative, entity alignment,

authority) and manage presence across the surfaces that shape answers (reviews, communities, publications, profiles), weighting toward the off-site 85%.

1. Scale + maintain. Run the operating cadence (monthly recon, biweekly answer pages,

4+ channel distribution) and set up attribution, because GEO shows a 4-8 week lag and mostly hides in "direct".

Prioritize high-intent queries (comparisons, "best for X", alternatives) first everywhere.

The plays

Audit / diagnose

• Master method: Diagnose -> Position -> Build -> Scale across on-page, authority, technical

entity, prompt mapping, and monitoring workstreams

• Run the AI-visibility audit: a query x model grid of unbranded buyer questions plus a

60-minute top-10-pages diagnostic, then classify gaps as coverage / consistency / specificity

On-page content structure (answerability)

• Structure content to get cited: 120-180-word sections, a self-contained 40-60-word answer

first, comparison tables, statistics, human-written

Off-site authority & entity consistency

• Earn off-site citations: get into listicles, comparison roundups, Reddit/Quora, and the

consideration set (85% of mentions are third-party)

• Build entity confidence: one identical positioning across site/LinkedIn/G2/Crunchbase/

Wikidata/Reddit, plus ecosystem presence on YouTube/Reddit

Operating cadence & attribution

• Run GEO as a recurring loop: monthly recon, biweekly answer pages (comparison -> use-case

-> topical authority), multi-channel distribution, and GA4 attribution for hidden AI traffic

Key numbers & benchmarks

• 85% of AI-answer brand mentions come from off-site third parties; 90% come from

listicles/comparisons plus Reddit/Quora. Comparative listicles are 32.5% of AI citations (listicles overall ~40%); for pro services, 80.9% of cited list content came from neutral third parties.

• Consideration set: cited 53.1% of the time when in it vs 10.6% when not.
• Model divergence: only ~11% domain overlap between ChatGPT and Perplexity, so test

5-10 unbranded buyer questions across ChatGPT / Perplexity / Gemini / Claude; re-run 15-20 questions monthly (~2-3 hrs).

• On-page citation lifts: 120-180-word sections get ~70% more citations; tables ~2.5x;

adding statistics ~41%; lead each section with a 40-60-word self-contained answer; 82% of AI-cited articles are human-written.

• Ecosystem correlation: YouTube presence correlates 0.737 with AI recommendations,

brand web mentions 0.664. Content across 4+ channels is 2.8x more likely to appear in ChatGPT. Baseline with a free AEO grader (0-100).

• Attribution: GA4 custom-channel regex chatgpt.com|perplexity.ai|gemini.google.com|copilot.com|claude.ai;

expect a 4-8 week publish-to-citation lag; ~70% of AI visitors arrive with no referrer (look like direct). AI-referred visitors convert ~11x organic; AI mentions convert 7.1% (vs 7.8% paid search).

• Classic SEO still feeds AI: a #1 Google ranking is ChatGPT-cited 43.2% of the

time; inline brand-homepage links in ChatGPT answers jumped 0.4% -> 6.2% (May 7, 2026).

• Cases: one fiber provider went 0% -> 40% appearance in 90 days; one retailer got

3.2x AI-referred traffic in 60 days via a documentation rewrite. AI search is projected to eclipse traditional search by 2028.

Reference library

Every play above, in full.

Get recommended by AI search (ChatGPT, Perplexity, AI Overviews) with the Diagnose → Position → Build → Scale method, because most of what AI says about you comes from off-site

The strategy

AI search optimization (GEO) is mostly a brand-positioning problem, not a technical one: the headline stat is that roughly 85% of what AI says about you comes from outside your website. You influence AI answers by engineering signals across the third-party sources models trust, tracked continuously. The method is four phases: Diagnose → Position → Build → Scale.

When to use it

When buyers in your category are asking AI tools (ChatGPT, Perplexity, Google AI Overviews) for recommendations and you want to be named, especially for high-intent, pipeline-driving prompts.

How to execute (steps)

1. Diagnose. Run structured testing across the major models to map current visibility, what drives it, and where the openings are. Identify your positioning vs competitors.
2. Position. Engineer the signals models read: narrative, entity alignment, and authority cues across trusted sources.
3. Build. Manage presence across the surfaces that shape answers, reviews, communities, publications, profiles, focusing on the sources models weigh most heavily.
4. Scale. Expand the footprint across organic and paid placements using live data.
5. Supporting workstreams:

• Content production, build comparison pages, category explainers, and proof content, structured the way models read and cite them.
• Authority & citations, earn mentions/reviews/citations on third-party sites and communities.
• Technical authority, align entities, schema, and structured data so models are confident about who you are and your category fit.
• Intent & prompt mapping, map the real prompts/buying questions customers pose to AI, and build coverage against the high-pipeline ones.
• AI visibility monitoring, track appearance across every major LLM continuously; adjust proactively before slipping out of answers.

Notes / caveats / examples

• Core insight to internalize: 85% of what AI says about you is off-site → prioritize third-party reviews, community presence, and citations over on-page tweaks.
• Comparison pages, category explainers, and proof content are the content formats that get cited most.

→ Skill conversion note

Becomes a "GEO / AI-search visibility" SKILL.md: runs prompt-set diagnostics across LLMs, maps buying prompts, and outputs a Position/Build/Scale action list weighted toward off-site authority sources.

Run an AI-visibility audit across every model, then diagnose your gaps as coverage, consistency, or specificity

The strategy

Before you can improve AI-answer visibility you have to measure it, per-model, because the models disagree sharply about who to recommend. The audit is a query×model grid of unbranded buyer questions; the diagnosis sorts every gap into one of three fixable types. This is the concrete "Diagnose" step under ai-search-optimization.

When to use it

• You want a baseline of how ChatGPT/Perplexity/Gemini/Claude describe and recommend you today.
• You suspect AI tools cite your content but recommend a competitor, and want proof.
• You need a prioritized list of which queries and gaps to fix first.

How to execute (steps)

1. Build a query×model grid. Query ChatGPT, Perplexity, Gemini, and Claude with 5-10 unbranded buyer questions and test all major models, there's only ~11% domain overlap between ChatGPT and Perplexity, so one model isn't a proxy for the rest.
2. Run the 60-minute page-based diagnostic. Take your top 10 organic pages, turn each into an AI query, test it across ChatGPT/Claude/Gemini, and log three things: brand appearance, direct links, and "ghost citations" (your content is cited while a competitor is the one recommended). Repeat quarterly.
3. Check description accuracy. For each answer, analyze whether the AI's description of you matches your actual positioning/differentiator, being mentioned wrong is its own gap.
4. Sort every gap into one of three types:

• Coverage, competitors have multi-source reinforcement you lack.
• Consistency, your messaging is contradictory across sources.
• Specificity, you're described broadly while a rival owns a niche.

1. Prioritize high-intent queries. Fix comparisons, "best for X", and alternatives queries first, they're closest to a buying decision.

Notes / caveats / examples

• The 11% ChatGPT/Perplexity overlap is the reason to audit every model, not just the biggest one.
• "Ghost citations" are the most actionable finding, the content already ranks well enough to be cited; the fix is positioning/recommendation, not more content.
• Feeds directly into the fix cards: content-structure-for-ai-citations, off-site-and-listicle-citations, entity-consistency-and-ecosystem.

→ Skill conversion note

Becomes an ai-visibility-auditor: input = brand + top pages + competitor set; output = a filled query×model grid with brand-appearance/ghost-citation flags, a gap classified per query (coverage/consistency/specificity), and a ranked fix list weighted to high-intent queries.

Structure on-page content to get cited by LLMs, 120-180-word sections, self-contained answers, tables, and stats

The strategy

LLMs extract and cite passages, not pages. So the on-page work in GEO is formatting content into self-contained, extractable chunks the models can lift verbatim. Each structural choice below carries a hard citation-lift number; this card is the on-page "clarity" lever (the off-site lever lives in off-site-and-listicle-citations).

When to use it

• You have content that ranks in Google but doesn't get pulled into AI answers.
• You're rewriting docs/comparison/use-case pages specifically for AI extractability.
• You want format rules with evidence, not opinion.

How to execute (steps)

1. Write 120-180-word sections. Sections in that range get ~70% more LLM citations than longer or shorter ones.
2. Lead each section with a self-contained 40-60-word answer. Put the extractable answer first, so a model can quote it without needing surrounding context.
3. Use comparison/markdown tables. Tables earn ~2.5x citations, structure comparisons and specs as tables, not prose.
4. Include statistics. Adding statistics lifts visibility ~41%, cite concrete numbers models can quote.
5. Keep it human-written. 82% of AI-cited articles are human-written, don't ship raw AI copy expecting it to be cited.

Notes / caveats / examples

• Cases: Bluepeak Fiber went 0% → 40% appearance in relevant AI queries in 90 days; Home Nation got 3.2x AI-referred traffic in 60 days via a documentation rewrite.
• These are formatting rules layered on genuinely useful content, a well-structured thin page still won't be trusted (see entity-consistency-and-ecosystem).
• The 40-60-word lead-answer pattern doubles as featured-snippet/AI-Overview optimization for classic search.

→ Skill conversion note

Becomes a citability-formatter: input = a draft/page; output = the page re-chunked into 120-180-word sections, each opening with a 40-60-word self-contained answer, comparisons converted to tables, and a flag on missing statistics.

Build AI's confidence in you, one consistent entity narrative across every profile, plus ecosystem presence on YouTube/Reddit

The strategy

Models recommend brands they're confident about, and confidence comes from a single, consistent story repeated across the surfaces they trained on. So GEO has an entity layer: score your baseline, make your positioning identical everywhere (site, LinkedIn, G2, Crunchbase, Wikidata), and seed authentic mentions on the platforms that correlate most with AI recommendations. This is the "consistency" fix from the audit, plus the ecosystem-presence fix.

When to use it

• Your audit found consistency or ecosystem gaps (contradictory messaging, thin third-party presence).
• Different profiles/pages describe you differently and you suspect it confuses the models.
• You want the durable, compounding GEO work rather than one-off content.

How to execute (steps)

1. Score an AEO baseline. Use HubSpot's free AEO Grader (0-100) to benchmark before you change anything.
2. Make your entity story identical everywhere. Repeat the same positioning across your site, LinkedIn, G2, Crunchbase, GMB, Wikidata (and Reddit) so LLMs gain confidence about who you are and your category. Consistency is the signal.
3. Fix the three AI gaps together: entity clarity (consistent positioning across site/LinkedIn/Crunchbase/G2/Reddit), answerability (repurpose existing SEO into extractable use-case/comparison pages), and ecosystem presence (be where models look).
4. Prioritize the highest-correlation ecosystem surfaces. YouTube presence correlates 0.737 with AI recommendations and brand web mentions 0.664, invest in video and earned mentions, not only your blog.
5. Seed authentic mentions (Loop Marketing). Run the Express / Tailor / Amplify / Evolve loop and seed genuine mentions on LLM-trained surfaces (YouTube, Reddit, newsletters) via media kits sent to creators.

Notes / caveats / examples

• Supporting stats: AI-referred leads convert 3-4x better; 60% of Google searches are zero-click; AI search is projected to eclipse traditional search by 2028, and citation rate is 53.1% when you're in the consideration set vs 10.6% when not.
• Entity consistency is unglamorous but compounding, every profile you align raises model confidence for every future query.
• Complements off-site-and-listicle-citations (earn the mentions), this card makes sure those mentions all tell the same story.

→ Skill conversion note

Becomes an entity-consistency-auditor: input = your profiles (site/LinkedIn/G2/Crunchbase/Wikidata/Reddit) + positioning; output = an AEO baseline score, a diff of inconsistent descriptions to fix, and a ranked ecosystem-seeding plan weighted to YouTube/Reddit/newsletters.

Run GEO as an operating cadence, monthly recon, biweekly answer pages, multi-channel distribution, and GA4 attribution

The strategy

GEO isn't a project, it's a recurring loop: monitor how the models describe you, publish answer pages on a steady beat, distribute across the surfaces models learn from, and measure the AI-referred traffic that mostly hides as "direct" in GA4. This card is the ongoing operating rhythm plus the attribution setup to prove it.

When to use it

• You've done the initial audit and fixes and need a sustainable monthly/biweekly process.
• Your analytics can't see AI-referred traffic, so GEO looks like it's doing nothing.
• You want to know which tools and what publishing cadence to commit to.

How to execute (steps)

1. Monthly recon (2-3 hrs). Re-run 15-20 buyer questions across ChatGPT / Perplexity / Gemini / Claude. Tools: Otterly AI, Peec AI, SE Ranking Visible.
2. Biweekly publishing. Ship 1-2 answer pages every two weeks in this order: comparison → use-case → topical authority.
3. Distribute across 4+ channels. Content distributed across 4+ channels is 2.8x more likely to appear in ChatGPT, don't just publish and wait.
4. Set up GA4 attribution. Add a custom-channel regex: chatgpt.com|perplexity.ai|gemini.google.com|copilot.com|claude.ai. Expect a 4-8 week lag between publishing and citation.
5. Account for hidden AI traffic. AI-referred visitors convert ~11x organic, but ~70% arrive with no referrer and look like "direct", the regex + first-party tracking is how you stop under-crediting GEO.
6. Mine ChatGPT Ads impression data as content research. Strong-delivering Context Hints (from the paid channel) reveal high-volume buyer problem-states to build organic content around, sequence paid vs organic by your current AEO maturity.

Notes / caveats / examples

• The 4-8 week lag is critical for expectations, GEO investments don't show for a month or two; judging weekly kills the program.
• The comparison → use-case → topical-authority publishing order builds citation surface fastest (comparisons and use-cases are the most-cited formats).
• Conversion stat to defend the program internally: 11x organic conversion on AI-referred visitors, most invisible in default GA4.

→ Skill conversion note

Becomes a geo-cadence-runner: input = buyer question set + content backlog; output = a monthly recon checklist (queries × models × tool), a biweekly publishing calendar in the right format order, a GA4 custom-channel regex, and a distribution plan hitting 4+ channels.

Earn AI citations off-site, get into listicles, comparison roundups, and the consideration set, because 85% of AI mentions come from third parties

The strategy

The single biggest GEO lever is off your own site: models mostly repeat what third parties say about you. So the work is getting named in the listicles, comparison roundups, and community answers that AI models pull from, and getting into the "consideration set," because being in it more than doubles your citation rate.

When to use it

• Your on-page content is already well-structured but you're still absent from AI answers.
• You want the highest-leverage GEO work, not incremental on-page tweaks.
• You're deciding where to spend PR/content effort for AI visibility.

How to execute (steps)

1. Prioritize off-site presence. 85% of AI-answer brand mentions come from third-party sources; 90% come from listicles/comparisons plus Reddit/Quora. Pursue listicles, review roundups, Reddit/Quora answers, and PR, not just your own blog.
2. Get into comparison listicles specifically. Comparative listicles are 32.5% of AI citations (and listicles overall capture ~40%). For professional services, 80.9% of cited list content came from neutral third parties, so aim for independent roundups, not your own comparison pages alone.
3. Build your own vs-pages too. Publish comparison/alternatives pages the way ClickUp built vs-Asana / vs-Monday / vs-Notion / vs-Jira pages, they capture "vs" and "alternatives" queries.
4. Get into the consideration set. Brands mentioned in the consideration set are cited 53.1% of the time vs 10.6% when not, being named at all, even alongside rivals, dramatically raises your odds.
5. Exploit the Branded Link Update (May 7, 2026). Inline brand-homepage links in ChatGPT answers jumped 0.4% → 6.2%; a #1 Google ranking gets ChatGPT-cited 43.2% of the time (3.5x pages beyond the top 20). Keep classic SEO strong, it feeds AI citation.

Notes / caveats / examples

• Conversion proof: organic AI mentions now convert 7.1% (vs 7.8% paid search); ChatGPT referral traffic is up ~60-65% and B2B SaaS referrals +200%. Case studies: Bluepeak Fiber 0% → 40% in 90 days; Home Nation 3.2x AI-referred traffic in 60 days.
• The neutral-third-party finding (80.9% for pro services) means earned placements beat self-published comparisons for credibility with models.
• Pairs with entity-consistency-and-ecosystem (be consistent across those third-party surfaces) and content-structure-for-ai-citations (make what you publish extractable).

→ Skill conversion note

Becomes an off-site-citation-planner: input = category + target queries; output = a target list of listicles/roundups/communities to get into, a vs-page build plan, and a consideration-set checklist ranked by citation-share potential.

From God of Skills: a curated, hand-tested directory of AI skills, prompts, templates and image style guides. Source: https://godofskills.com/skills/ai-search-geo?ref=claude-skill