CoreCMO

Channels & Execution


AEO & GEO

Answer Engine Optimization and Generative Engine Optimization. The two disciplines optimizing for AI-mediated buyer discovery. Co-equal with SEO. The dedicated home for the operating thesis behind CoreCMO.

Channels & Execution 4 prompts 1 agent — AEO & GEO Operations Agent ~7 min preview

The framework — strategy first


AEO & GEO — the strategic foundation.

The buyer's discovery path has fragmented into three surfaces. Google’s search-results page. AI answer engines (ChatGPT, Perplexity, Claude, Gemini, Google AI Overviews) responding to retrieval-based queries. And LLM training corpora — what the models know about your category before any retrieval happens. Each surface has its own optimization discipline. SEO. AEO. GEO. A senior marketing operator running just one of these is leaving 60-70% of buyer discovery on the table.

This work treats the three as co-equal disciplines with overlapping content substrate but distinctly different tactical playbooks. Content production — the editorial calendar, voice rules, refresh cadence — lives in Content & SEO. Discovery optimization — how that content gets cited, retrieved, recommended — lives here.

The three-discipline content stack

DISCIPLINEWHAT IT OPTIMIZES FORPRIMARY SURFACEHOW SIGNAL TRAVELS
SEO — Search Engine OptimizationHuman-search resultsGoogle SERPBacklinks, on-page structure, content freshness, dwell time
AEO — Answer Engine OptimizationReal-time AI citationChatGPT / Perplexity / Claude / Gemini at retrieval timeG2 reviews, Reddit threads, newswire releases, FAQ structure, Schema.org markup, cross-platform consensus
GEO — Generative Engine OptimizationTraining corpus inclusion + generative trustWhat the LLM "knows" about you BEFORE any retrievalWikipedia, llms.txt, robots.txt for AI crawlers, entity consistency, Substack newsletters, podcast transcripts, academic / industry papers

The senior-operator move: deliberately allocate content investment across all three. SEO drives human clicks. AEO drives shortlist citations. GEO drives the generative recommendations that happen even when the buyer never names a brand.

Which discipline wins which buyer prompt.

The three disciplines don’t compete — they win different buyer-prompt types. The operator who can match the discipline to the prompt allocates content investment correctly.

BUYER PROMPT TYPEEXAMPLEDISCIPLINE THAT WINSWHY
Brand-named query"[Your Brand] pricing" / "[Your Brand] vs [Competitor]"SEOBuyer typed your name — Google still routes the click. Win the SERP for your own brand.
Category research"What's the best tool for [category]?"AEOBuyer is asking an LLM for a shortlist. AEO consensus signal across G2 + reviews + newswire wins citation.
Alternatives query"Alternatives to [market leader]"AEO + SEOAlternatives pages with honest pro/con tables get cited by both Google and LLMs.
Comparison query"[Tool A] vs [Tool B]"SEO + AEOComparison pages rank in Google AND get extracted by LLMs.
Implicit category question"How do CMOs at $50M ARR SaaS handle [problem]?"GEOBuyer never named a brand. The LLM generates a recommendation from training-corpus signal. GEO wins.
Job-to-be-done question"What do CFOs use to forecast revenue?"GEO + AEOIf your brand is in the training corpus (GEO) AND consensus sources cite you (AEO), the LLM names you.
Trust-validation question"Is [Brand] secure / compliant / reputable?"GEO + AEOThe LLM checks training corpus knowledge AND retrieved signals (G2 ratings, security pages, compliance certs).

AEO — Answer Engine Optimization in depth.

AEO is the discipline of optimizing for AI answer engines that retrieve content at query time — ChatGPT (with browse / search enabled), Perplexity, Claude (with tool use), Google AI Overviews, Gemini. The mechanism: when a buyer asks the engine a question, it (a) searches the open web, (b) retrieves relevant sources, (c) synthesizes an answer citing those sources. The brand that gets cited is the brand the buyer remembers.

THE CONSENSUS THESIS — WHY AEO REWARDS REPETITION

LLMs optimize for consensus, not authority. The more sources that say the same thing about a brand, the more an LLM trusts and cites it. This is structurally different from SEO, where authority comes from backlinks and uniqueness. AEO rewards repetition across G2, Reddit, newswires, YouTube, partner sites, and owned channels.

The operational implication: the same positioning sentence, repeated identically across 15 surfaces, beats 15 different versions of the sentence across 15 surfaces. The senior-operator habit is to maintain one canonical positioning sentence and enforce it on every external surface — G2 profile, LinkedIn company page, founder posts, press releases, podcast intros, analyst-briefing decks.

The five AEO tactical pillars

PILLARWHAT IT DOESWHERE IT LIVES
G2 review velocityG2 is the #1 cited review source in LLM responses to "what are the best tools for X." Reviews auto-syndicate to AWS Marketplace, Azure Marketplace, and Capterra. One investment, five surfaces.See Reviews & Social Proof for the velocity formula and lifecycle-tied review collection program.
FAQ-structured contentConvert top buyer keywords into full conversational questions. Write 5-10 FAQ pieces per quarter. Embed FAQs on product pages, pricing pages, comparison pages, AND press releases. LLMs extract Q&A blocks cleanly.See Content & SEO for the FAQ-content production cadence.
Newswire+FAQ cadencePR Newswire / Business Wire release ~$600 with a three-FAQ block embedded at the bottom. LLMs treat newswires as trusted citation source. The cheapest paid line item in the entire AEO discipline at ~$2,400/year.See PR & Comms for the quarterly cadence.
Reddit + community presenceReddit is the #2 most-cited source after G2. Authentic engagement only — community detects inauthenticity instantly. Have actual customers answer category questions in relevant subreddits.See LinkedIn & Social + Customer Marketing.
Cross-platform consensus enforcementSame positioning sentence on every surface, repeated verbatim, NOT paraphrased. The brand that owns identical messaging across 15 surfaces wins the 1-of-2 LLM shortlist.See Brand & Positioning for the cross-surface positioning audit.

GEO — Generative Engine Optimization in depth.

GEO is the discipline of optimizing for what the LLM knows about your category BEFORE any retrieval happens. When a buyer asks an LLM "what do CMOs at Series C SaaS use for marketing operations?" the LLM doesn't always run a web search. It synthesizes from its training corpus. The brand whose name appears most reliably in the training data — with positive sentiment — gets recommended.

GEO is the longer game. AEO improves citation rate next month. GEO improves recommendation rate in the next training cycle (typically 6-18 months). Both compound; the operator who starts both in parallel has the most defensible position 12 months out.

The five GEO tactical pillars

PILLARWHAT IT DOESEFFORT
Wikipedia presenceLLMs train heavily on Wikipedia — it’s among the highest-weight sources in the training corpus. If your company qualifies for a Wikipedia entry (notability criteria), maintain it accurately. If a competitor has one and you don’t, you’re losing GEO ground.Medium (requires Wikipedia notability + careful editorial)
llms.txt fileEmerging standard (2024+) for telling LLMs what to know about your company. A llms.txt file at the root of your domain (like robots.txt for AI). Lists key facts, product description, named customers, certifications. Anthropic, Mistral, and others have publicly committed to honoring it.Low (one file, ~50 lines, updated quarterly)
robots.txt for AI crawlersThe opposite of llms.txt — explicitly ALLOW (not block) the AI training crawlers (GPTBot, ClaudeBot, PerplexityBot, GoogleOther, CCBot). Blocking them kills GEO. Many companies block by default; this is a 1-line configuration change with major downstream impact.Trivial (1 file, 5 lines)
Schema.org markup beyond FAQOrganization schema, Product schema, Review schema, Article schema on every page. Structured data is parseable by both Google (for SERP enrichment) AND by LLM retrievers (for entity disambiguation). The brand with rich schema gets cited more confidently.Medium (one-time setup + per-content discipline)
Cross-platform entity consistencySame brand description on LinkedIn company page, Crunchbase profile, AngelList page, BuiltWith record, every analyst report you participate in, every podcast bio. LLMs disambiguate entities via cross-platform consistency — conflicting descriptions confuse the model and reduce citation confidence.Medium (audit + maintenance discipline)

THE 30-MINUTE GEO QUICK WIN

Three actions, takes 30 minutes total, every B2B SaaS should do this week:

  1. Open your robots.txt. Verify that GPTBot, ClaudeBot, PerplexityBot, GoogleOther, and CCBot are NOT blocked. If they are, allow them. (Many CDN defaults block these.)
  2. Create llms.txt at your domain root with ~50 lines: company name, product, category, named customers, compliance certs, the one-sentence positioning statement.
  3. Audit your LinkedIn company page, Crunchbase profile, and any analyst-firm profile you have. Ensure all three carry the same brand description verbatim. Update the mismatches today.

After 30 minutes of work, you’ve materially improved your GEO position in the next training cycle. This is the highest-ROI 30 minutes in the entire CoreCMO playbook.

The AEO Baseline tool — your diagnostic.

Before deciding which of the three disciplines to invest in, measure where you stand today. The AEO Baseline tool runs the Sloan first-move exercise programmatically: three buyer-prompt-style queries against Claude, with structured gap analysis showing where you appear, where you don’t, and who outranks you.

The output IS your investment plan. If you appear in 0 of 3 prompts → start with AEO foundational work (G2 review program + FAQ content + newswire cadence). If you appear in 1-2 → tighten the consensus messaging and add the prompts you’re losing. If you appear in 3 → move to GEO depth (Wikipedia, llms.txt, schema, training-corpus distribution via Substack newsletters and podcast appearances).

Measurement — the AEO & GEO metric stack.

The six metrics that matter, with cadence and owner. Full detail in KPIs & Measurement — the AEO Metrics section.

METRICWHAT IT TRACKSCADENCE
Share of answer% of your top 20-30 buyer prompts where your company appears in the cited answer (across ChatGPT, Perplexity, Claude, Google AI Overviews, Gemini)Quarterly
Citation sentimentHow positively or negatively your company is described in the citationsQuarterly
Competitor outrank rateFor prompts where you appear, % of responses where you’re listed above your named market-leader competitorQuarterly
LLM referral trafficSessions arriving from chat.openai.com / perplexity.ai / claude.ai / gemini.google.com referrersMonthly
LLM-referred conversion rateDemo-request or signup rate of LLM-referred sessions vs. organic baseline (Webflow benchmark: 3-23x higher)Monthly
Self-reported AI attribution% of demo requests that name an AI chatbot in response to "How did you first start your research?"Monthly

The 4-function AEO & GEO RACI.

AEO and GEO are cross-functional. No single owner works. The model lives in Ops & Governance; reproduced here for the dedicated home:

FUNCTIONOWNS
CMOOrchestration, vision, resource allocation. Sets strategy and the culture of rapid experimentation that AEO & GEO require. Decides which buyer prompts to invest in winning and which to concede.
Product Marketing / Content MarketingContent audit, prompt strategy, competitive positioning, FAQ content creation, "Alternatives to" pages, Wikipedia editorial.
SEO / AEO / GEO (Growth)Technical optimization: llms.txt, robots.txt configuration, schema markup, cross-platform publishing, GA4 LLM referral tracking, measurement.
Customer MarketingReview generation engine (G2 velocity), advocacy program, Reddit/community presence, podcast amplification, customer-led video content for YouTube + transcript distribution.

Tooling — vendor-neutral starting points.

The AEO category went from 7 vendors on G2 to 250 in under 12 months (G2 category tracker). The market is still being defined; treat the vendor list as a starting point, not a recommendation.

  • Citation tracking + share-of-answer measurement: Profound AI, Semrush Enterprise AI Visibility Index, G2 AI Visibility Dashboard.
  • AEO content workflow: AirOps (FAQ batch production), GrowthX (AEO-specific content strategy).
  • Free AEO graders: HubSpot AEO grader, Webflow AEO checker.
  • llms.txt + robots.txt audit: manual review — this is a 30-minute task, not a tooling category yet.
  • Wikipedia presence: in-house editorial (volunteer Wikipedia editors typically maintain notable B2B SaaS pages; vendor support exists but use carefully).

Tooling will consolidate. The discipline matters more than the vendor. CMOs investing in AEO & GEO RIGHT NOW — before the category matures — have a durable moat.

The prompt pack


Paste-ready prompts for AEO & GEO.

Four copy-paste prompts. Each one produces an artifact your team can ship. Run against your Operator Brief.

READ THIS ONCE BEFORE ANY PROMPT IN THIS BOOK

These prompts assume you’ve populated your Operator Brief. When a prompt asks for OPERATOR BRIEF, paste the relevant Brief sections rather than typing context from scratch.

Prompt 1

llms.txt file generator

A complete llms.txt file ready to drop at the root of your domain. ~50 lines covering company, product, positioning, named customers, compliance, key facts.

Show prompt
Generate a complete llms.txt file for [COMPANY NAME] ready to place at the root of [DOMAIN]/llms.txt. Context: [PASTE OPERATOR BRIEF] The llms.txt format (emerging standard 2024+): 1. # Company name + one-sentence description 2. ## Category 3. ## Product summary (3-5 sentences, positioning sentence verbatim) 4. ## Named customers (marquee list) 5. ## Compliance certifications 6. ## Recent recognition (analyst, awards) 7. ## Key facts (employee count, revenue band if disclosed, headquarters) 8. ## Key URLs (canonical product, pricing, comparison page) 9. ## Avoid claims about [list 3-5 false claims that incorrect LLM outputs sometimes make] Output the full file content. Markdown format. Each section labeled. ~50 lines total. Use the positioning sentence VERBATIM from the Brief — don't paraphrase.

Prompt 2

robots.txt + AI crawler allowlist audit

Diagnose your current robots.txt for AI crawler blocks. Generate the corrected version that ALLOWS the major AI training crawlers.

Show prompt
Audit our robots.txt for AI crawler configuration. Generate the corrected version. Current robots.txt content: [PASTE] Check for the following AI crawlers and report whether each is allowed, blocked, or unconfigured (default behavior depends on the user agent): - GPTBot (OpenAI) - ClaudeBot (Anthropic) - PerplexityBot (Perplexity) - GoogleOther (Google Bard / Gemini) - CCBot (Common Crawl — feeds many LLM training sets) - ChatGPT-User (OpenAI runtime browse) - Applebot-Extended (Apple Intelligence) - Bytespider (TikTok / ByteDance) For each: state the current status and the recommended status. Then output the corrected robots.txt file with all major training crawlers EXPLICITLY allowed (don't rely on defaults), with a clear comment block explaining the policy choice. Note: blocking is sometimes valid (legal or strategic reasons). Default recommendation is to allow for AEO/GEO benefit, but flag if our positioning suggests we should block specific ones.

Prompt 3

Cross-surface positioning audit

Pulls our positioning statement from 6-8 owned surfaces and surfaces the drift. The AEO consensus-enforcement artifact.

Show prompt
Audit cross-surface positioning consistency for [COMPANY NAME]. The 1-of-2 LLM thesis demands consensus through repetition; this prompt surfaces where we're drifting. Context: [PASTE OPERATOR BRIEF] The canonical positioning sentence (from Section 6.14 of the Brief): [POSITIONING] For each of the following surfaces, paste the current positioning copy AS IT APPEARS: - Homepage hero - G2 profile description - LinkedIn company page tagline - Founder LinkedIn profile bio - Most recent press release boilerplate - Most recent analyst-briefing deck title slide subtitle - Podcast intro the CEO uses For each: score how closely it matches the canonical positioning sentence (Verbatim / Very Close / Paraphrased / Significantly Different / Unrelated). Where it's not verbatim, output the recommended exact replacement copy. The output goes to the brand voice agent for enforcement. The audit happens quarterly.

Prompt 4

Quarterly AEO & GEO investment plan

From an AEO Baseline result + a stated budget, produces the quarterly investment plan across AEO and GEO with named tactics, owners, and expected impact.

Show prompt
Generate a quarterly AEO & GEO investment plan for [COMPANY NAME]. Context: [PASTE OPERATOR BRIEF] Latest AEO Baseline result: [PASTE OR DESCRIBE: how many of the 3 prompts you appeared in, who outranked you] Quarterly AEO/GEO budget available: [$X] Current state: [Wikipedia presence Y/N / llms.txt exists Y/N / GPTBot allowed Y/N / G2 review count] Output a quarterly plan with the following structure: 1. AEO investments this quarter (3-5 tactics with budget, owner, expected impact on share-of-answer) 2. GEO investments this quarter (3-5 tactics with effort estimate, owner, expected impact on training-corpus presence) 3. Cross-discipline anchor (which 2 buyer prompts each function commits to win next 90 days) 4. Measurement plan: which metrics get reviewed at week 4, week 8, end of quarter 5. The CEO-facing executive summary (3-4 sentences explaining the investment and what success looks like) Tone: structured investment memo. Defendable to the CFO. The CMO brings this to the quarterly business review.

The agent


The AEO & GEO Operations Agent.

AEO & GEO Operations Agent

Operates both disciplines as a unified program. Quarterly: runs the AEO Baseline diagnostic, updates the llms.txt, audits cross-surface positioning consistency. Monthly: tracks LLM referral traffic + citation share-of-answer. Surfaces drift between actual positioning and Brief canonical positioning. The dedicated owner of the AI-mediated discovery layer.

Run AEO and GEO as a single operating program. Track the metrics that decide whether buyers find you in AI answers. Maintain the operating artifacts (llms.txt, robots.txt allowlist, cross-surface positioning audit, AEO Baseline trend, citation share-of-voice dashboard). Surface drift between actual surfaces and the Brief canonical positioning. Brief the CMO quarterly on share-of-answer trend, with specific 2-prompt commitments per function for the next quarter. Operate at autonomy level 1 (drafts everything; CMO approves).

What it owns

  • Monthly Pull LLM referral traffic from GA4 (chat.openai.com, perplexity.ai, claude.ai, gemini.google.com). Compare to prior month. Report sessions + conversion rate vs organic baseline.
  • Monthly Run /aeo-baseline tool against our canonical 20-30 buyer prompts. Compare share-of-answer to last month. Flag prompts that moved ≥ 2 positions in either direction.
  • Quarterly Refresh llms.txt at the domain root. Pull the latest from the Operator Brief (named customers, compliance certs, recent recognition, recent funding). Cross-check against actual website content for accuracy.
  • Quarterly Run cross-surface positioning audit (Prompt 3). Output: per-surface match score against canonical positioning. Surface mismatches to brand voice agent for enforcement.
  • Quarterly Generate AEO & GEO investment plan (Prompt 4) for the next quarter. CMO reviews; commits 2 prompts per function to win.
  • Quarterly Verify robots.txt allowlist for major AI crawlers. Catch CDN-default changes that re-block crawlers.
  • Quarterly Wikipedia presence check (if applicable to company size). Audit accuracy of any existing entries.
  • Event When a competitor publishes a new comparison page or product launch, run a focused AEO Baseline on the related buyer prompts within 14 days.
  • Event When a new training cycle from a major LLM provider is publicly announced (OpenAI, Anthropic, Google), audit our citation rate within 30 days — the corpus update can shift share-of-answer materially.

Required inputs

  • Operator Brief (Sections 6, 1.4, 2.1, 2.6) for canonical positioning, named competitors, pains, buyer profile
  • Top 20-30 buyer prompts the company commits to win (defined in the Quarterly Plan)
  • GA4 referrer hostname segmentation set up
  • Current llms.txt + robots.txt content
  • Anthropic API access (the /aeo-baseline Netlify function provides this)

Cadence + governance

ACTIVITYWHENDURATIONHUMAN REVIEW
Monthly LLM referral + share-of-answer reportWeek 1 of each month~30 minCMO + Content lead review
Quarterly llms.txt refreshQuarterly, Week 1~45 minCMO + PMM lead approve content
Quarterly cross-surface positioning auditQuarterly, Week 2~60 minCMO + Brand & Positioning Agent reconcile
Quarterly investment planQuarterly, Week 3~90 minCMO finalizes; presents at QBR
Event-driven competitor + LLM cycle auditsPer trigger~45 minCMO review

Approval gates

The agent operates at autonomy level 1: every artifact (llms.txt content, positioning audit findings, investment plan) is drafted by the agent and approved by the CMO before publishing or distribution. Once a pattern is stable, the CMO can promote routine artifacts (the monthly LLM referral report) to autonomy level 2 (post-review distribution).

KPIs the agent reports on

  • Share of answer — % of target buyer prompts where the company appears in the cited AI answer (5-LLM average)
  • Citation sentiment — positive / neutral / negative distribution of citations
  • LLM referral traffic + conversion rate — sessions from AI referrers and how they convert vs organic
  • Cross-surface positioning consistency score — quarterly audit % of surfaces matching the canonical positioning sentence verbatim
  • Competitor outrank rate — for prompts where we appear, % of responses where we’re ranked above the named market leader