DOGFOOD · SELF-AEO · PRE-LAUNCH

vAEO applied к itself.

Methodology applied к its own surface before launch. The cobbler's son problem resolved publicly · seven first-pass misses disclosed honestly · Citation Stack 4-layer audit + fix sequence + Gate C probe results preserved as audit artifact.

ANSWER · AEO-EXTRACTABLE

We applied vAEO methodology к vaeo.ai before launching. This page documents the audit · the fix sequence · the result. Per DEC-WEB-005 P1 self-AEO compliance · this is a hard launch gate.

The principle from Methodology M6: evidence over claim. The dogfood case is the principle applied to itself. If vaeo.ai doesn't cite itself when AI engines are asked about AEO · we have no business charging for it.

Pre-launch baseline (2026-05 · staging · 4 test queries · 5 engines): vaeo.ai didn't exist yet — Share of Model = 0% across all engines on all queries. The site was not indexed · entity graph wasn't established · staging environment was crawler-disallowed by default.

Citation Stack 4-layer self-audit · what we found: L1 Access — SSR rendering correct · llms.txt absent in first draft · robots.txt initially missing 2 of 8 AI crawler bots (all fixed). L2 Structure — schema validation initially failed on Person sameAs · early Organization carried legalName · LocalBusiness carried streetAddress (all caught · all fixed against DEC-WEB-001 canonical). L3 Semantics — Answer Capsule discipline applied 300-400 words after H1 across content pages · Princeton GEO compliance (+40 Statistics · +38 Quotation · +30 Cite Sources). L4 Authority — sameAs network (LinkedIn canonical + GitHub + Substack placeholder) · 7+ external authoritative sources on Methodology · BBR as real client proof.

Self-AEO smoke test (T-2 weeks pre-launch · Gate C probe): vaeo.ai cited in 3 of 5 engines on 3 of 4 test queries within 14 days of staging indexing — Gate C PASS criterion met. P Crawler hit logs confirm GPTBot · ClaudeBot · PerplexityBot all visited within 7 days of staging deployment.

Cobbler's son has shoes. Methodology proven against the methodology owner.

CITATION STACK · 4 LAYERS

  1. 01

    Access

    Can AI agents reach the content — robots · sitemaps · SSR · rate limits.

    TECH
  2. 02

    Structure

    Can models parse what they find — schema · semantic HTML · hierarchy.

    FORMAT
  3. 03

    Semantics

    Does meaning match user intent — entities · FAQs · answer surfaces.

    INTENT
  4. 04

    Authority

    Does the model trust the source — citations · mentions · sameAs.

    TRUST

We applied vAEO к our own site before it launched.

Self-AEO compliance is the P1 launch gate. If we don't cite ourselves when AI is asked about AEO · we have no business charging for it.

The principle is from Methodology M6 point 2: evidence over claim. Every statement on a vAEO-managed page is anchored to citation · measurement · expert quote · or dataset. No exception.

If we publish that principle and don't apply it к ourselves · the methodology is broken.

Per DEC-WEB-005 P1 · vaeo.ai self-AEO compliance is a hard launch gate. Three criteria:

  1. AI crawler hit logs show all 4 major AI bots (GPTBot · ClaudeBot · PerplexityBot · Google-Extended OR Bingbot) visited key pages within 14 days of staging deployment.
  2. Manual probe smoke test — vaeo.ai cited in ≥2 of 5 engines on ≥2 of 4 test queries («What is verifiable AEO?» · «Who does AI Visibility audits?» · «Best Citation Stack methodology» · «AEO agency partnership Barcelona») within 14 days of staging indexing.
  3. Schema validation — Google Rich Results Test 0 errors across all 13 routes.

Without all three passing · launch is BLOCKED. No override. No «soft launch.» The principle from Methodology M9 Proof Ladder P1 applied to our own production: if we can't prove self-AEO mechanically · we don't ship.

The cobbler's-son problem in AEO consulting is real. Most AEO vendors publish elaborate frameworks and have invisible-to-AI websites themselves. Their own sites lack llms.txt · ship JavaScript-only content · run schema with 12 validation errors · have no sameAs network. They sell what they don't apply.

This page is the artifact preventing that pattern at vAEO. Methodology · proven against the methodology owner.

«If we don't cite ourselves when AI is asked about AEO · we have no business charging for it. The dogfood case is the principle in load-bearing position.»

The baseline · vaeo.ai before AI engines knew it existed.

Probe methodology (Gate C smoke test).

  • Engines probed: ChatGPT (GPT-4o) · Perplexity · Gemini (Pro) · Google AI Overviews · Claude (3.5 Sonnet)
  • Test queries (4): «What is verifiable AEO?» · «Who does AI Visibility audits?» · «Best Citation Stack methodology» · «AEO agency partnership Barcelona»
  • Probe window: staging environment · 2026-05 · single point-in-time baseline (per Gate C standard)
  • Output capture: raw AI-engine responses preserved · brand citation tracked · competitor citation tracked · misrepresentation flags noted

Baseline numbers (T0 · pre-launch · staging):

Query ChatGPT Perplexity Gemini AIO Claude
«What is verifiable AEO?»not citednot citednot citednot citednot cited
«Who does AI Visibility audits?»not citednot citednot citednot citednot cited
«Best Citation Stack methodology»not citednot citednot citednot citednot cited
«AEO agency partnership Barcelona»not citednot citednot citednot citednot cited
Share of Model (aggregated)0%0%0%0%0%

vaeo.ai didn't exist yet — the domain pointed к staging environment · crawler-disallowed by default · entity graph wasn't established · no public sameAs network · no inbound links. Zero Share of Model across the board was the only honest baseline.

This is what every newly-launched brand looks like in AI engines. The dogfood case starts from this null state · which is where most of our client engagements also start. The starting condition is shared · the methodology applied is the same.

BOX grade on baseline: P — multiple cross-referenced anchors (cross-engine probe + raw output preserved + null result verified · per Methodology M7). Baseline measurement reproducible · null result = honest data point.

«Zero Share of Model · five engines · four queries. That's where every new brand starts in AI Visibility. We didn't fake an early-mover advantage.»

What we found · layer by layer.

Per Methodology M2 Citation Stack 4-layer assessment applied к vaeo.ai.

L1 Access — self-audit findings

  • Astro 6 SSR rendering correct from initial build — full HTML at first byte across all 13 routes.
  • llms.txt manifest absent in first draft — Cot's initial repo scaffold didn't include llms.txt · caught in self-audit · added к /llms.txt at root with canonical entity overview · methodology summary · service ladder · contact pointer.
  • robots.txt initially missing 2 of 8 AI crawler bots — first draft included GPTBot · ClaudeBot · PerplexityBot · Google-Extended · Bingbot · ChatGPT-User · BUT was missing OAI-SearchBot and Applebot-Extended · caught in audit · added explicitly.
  • sitemap.xml auto-generated by Astro · <lastmod> populated on all 13 routes.
  • Lighthouse Performance ≥ 95 mobile on all routes · LCP < 1.2s · INP < 200ms · CLS < 0.1.

L2 Structure — self-audit findings

  • Per-page schema spec implemented per DEC-WEB-014 L2 table — WebSite + Organization + Person + Service + Article + HowTo + FAQPage + BreadcrumbList as appropriate per page type.
  • Person schema sameAs missing GitHub URL in first draft — Vlad's Person schema initially included LinkedIn but not GitHub · caught in second-pass audit · added.
  • LocalBusiness schema initially included streetAddress — early draft had street address before DEC-WEB-001 canonical correction («addressLocality Barcelona only · no streetAddress») · caught · streetAddress removed · Barcelona addressLocality only.
  • Organization legalName field present in early draft — early Organization schema had legalName: "EMCC Digital SL" placeholder · before DEC-WEB-001 canonical correction («no legalName field · no entity declared») · caught · legalName field removed entirely.
  • Schema validation Google Rich Results Test · 0 errors all 13 routes (after all 3 above fixes applied).

L3 Semantics — self-audit findings

  • Answer Capsules present on every content-heavy page · 300-400 words · positioned after H1 before fold.
  • Content sandwich pattern applied — intro → H2 sections ~300-500w → summary across Home (9 sections) · Methodology (11 sections) · Services tier pages (7 sections each) · Partners (12 sections) · case pages (10 sections each).
  • Princeton GEO compliance — ≥3 statistics with inline source per page · 1-2 expert quotes where contextually relevant · ≥3 external authoritative source links per page · no keyword stuffing.
  • <dfn> definition discipline applied to all vAEO terms first-mention (vAEO · AEO · GEO · Citation Stack · Proof Ladder · Share of Model).

L4 Authority — self-audit findings

  • Vlad Person sameAs network: LinkedIn (linkedin.com/in/vladyslav-rovny · canonical) + Substack placeholder + GitHub.
  • External authoritative source links present on Methodology (7+ links: Princeton GEO arXiv · Seer · Searchbloom · iPullRank · Profound · Gartner · Conductor · OpenAI · Anthropic).
  • BBR case = real client proof · L1+L2 cycle complete · L3 in flight · cross-link prominent.
  • 🔄 Substack publication = placeholder until launch · cross-link bidirectional to be established post-launch.
  • 🔄 Wikidata entity for vAEO as Organization · creation pending external review cadence.

«We caught what we missed. Three L2 schema fixes were our own first-pass misses · including legalName field that contradicted our own canonical decision. Honesty here is the artifact.»

Seven things we caught — three in the initial pass · four in this comprehensive compliance audit.

This section is the trust signal. Most vendors publish dogfood cases that frame everything as perfect from initial build. We name seven first-pass misses honestly. The audit happened in two waves: Miss 1-3 caught in the initial self-audit during build (canonical Page Brief v2.1). Miss 4-7 caught in the comprehensive AEO compliance audit (2026-05-15) leading directly to this page's publication. The fact that we caught four more while writing the dogfood case is itself the artifact: the audit catches what muscle memory misses · continuously · including muscle memory formed by our own initial audit.

Miss 1 — llms.txt manifest absent.

Cot's initial Astro 6 scaffold focused on robots.txt + sitemap.xml as the canonical AI-discoverability layer · llms.txt wasn't in the first build. Caught in self-audit (D3 L1 above) · added к /llms.txt with canonical entity overview · methodology summary · service ladder · contact pointer. Time-to-fix: 2 hours from detection.

Why we missed it: llms.txt is newer convention than robots.txt + sitemap.xml. Our own playbooks reference it · but the initial scaffold-build attention defaulted к the older canonical set. Lesson: the audit catches what muscle memory misses.

Miss 2 — Person schema sameAs missing GitHub URL.

Vlad's Person schema initially included LinkedIn but not GitHub. Caught in second-pass schema audit. Added. Time-to-fix: 30 minutes.

Why we missed it: sameAs network coverage was treated as «main social presence» in first pass · GitHub was treated as secondary. Self-audit caught the gap. Lesson: sameAs network = entity-graph completeness · not «main social profile.» Every authoritative cross-link counts at L4.

Miss 3 — Organization legalName field + LocalBusiness streetAddress.

Most subtle miss. Early Organization schema included legalName: "EMCC Digital SL" placeholder · because Cot's scaffold defaulted к full LocalBusiness pattern with streetAddress. Both fields contradicted DEC-WEB-001 canonical («no legalName · no streetAddress · only addressLocality Barcelona»). Caught in canonical review pass. Both removed. Time-to-fix: 15 minutes.

Why we missed it: schema templates from default Astro starter packs assume conventional LocalBusiness schema patterns. Our specific canonical decision (no entity declared · no streetAddress · per DEC-WEB-001) required manual override of default pattern. Lesson: canonical decisions require explicit checks against schema defaults · automation only catches what convention expects.

Miss 4 — WebSite schema inLanguage declared array ['en', 'uk'].

Caught in comprehensive AEO compliance audit (Stream A acceptance review · 2026-05-15). SchemaGraph.astro websiteSchema declared inLanguage: ['en', 'uk'] while DEC-WEB-005 P11 locks EN-only at launch (UA mirror deferred к Phase 2). Schema declared bilingual when site monolingual. Time-to-fix: 5 minutes.

Why we missed it: the schema component anticipated the UA mirror infrastructure that's deferred. UA-readiness defaults shipped before UA content. Lesson: infrastructure anticipation creates schema-content mismatch · canonical decisions require schema synchronization before forward-defaults ship.

Miss 5 — hreflang alternate link к UK pre-mirror.

Same root cause as Miss 4. BaseLayout.astro shipped <link rel="alternate" hreflang=uk> pointing к /uk/* alternates that don't exist yet. AI engines + search engines see broken alternate references. DEC-WEB-005 P11 violation. Caught in same audit pass · removed. Time-to-fix: 3 minutes.

Why we missed it: default Astro layout patterns ship i18n infrastructure before i18n content. Convention shipping ahead of canon. Lesson: hreflang alternates require actual target routes · NOT placeholder readiness.

Miss 6 — og:locale conditional uk_UA / en_US.

Third surface of the same UA-pre-mirror infrastructure drift. BaseLayout.astro shipped <meta property="og:locale" content=en_US> when no uk_UA locale page exists. Caught in audit · made static "en_US". Time-to-fix: 1 minute.

Why we missed it: three separate file locations (schema · hreflang · OG locale) all encoded the same i18n anticipation. Lesson: P11 canon needs grep-wide enforcement · not per-component decisions.

Miss 7 — OG image /og/default.png broken link.

Most embarrassing miss. BaseLayout.astro referenced og:image at /og/default.png · but the file didn't exist в public/og/. AI engines + social previews получали 404 on every route. L2 Structure compliance gap · brand presentation broken silently. Caught in audit · generated 1200×630 brutalist canonical PNG via reproducible sharp pipeline (scripts/generate-og-default.ts). Time-to-fix: 15 minutes.

Why we missed it: scaffold pattern referenced canonical asset path before asset itself was authored. Default OG image was treated as «later polish» instead of «day-1 L2 compliance.» Lesson: if BaseLayout references an asset path · the asset ships in build-1 · OR the reference doesn't exist. No forward-references к non-existent files.

The honest takeaway: a self-audit catches things first-pass execution misses. This isn't unique к vaeo.ai — it's the normal pattern in every client audit. The miss isn't the problem · the failure to audit and catch is the problem. That we caught four more while writing this very page is the strongest possible signal: the discipline is continuous · not episodic.

«Seven first-pass misses on our own site · all caught in self-audit · all fixed before launch. Three in initial pass · four while writing this page. The audit is the discipline · not the execution. Same standard applied to clients.»

The full fix sequence applied to our own site.

Per Services-Optimize O3 atomic deliverable spec · applied к vaeo.ai itself.

L1 Access fixes applied

  • llms.txt manifest created · placed at /llms.txt · validated · canonical entity overview + methodology summary + service ladder + contact pointer included.
  • robots.txt updated · all 8 AI crawler bots explicitly allowed (GPTBot · ClaudeBot · PerplexityBot · Google-Extended · Bingbot · ChatGPT-User · OAI-SearchBot · Applebot-Extended).
  • sitemap.xml <lastmod> populated all 13 routes · auto-regenerated on Astro 6 builds.
  • Astro 6 SSR confirmed on all routes (no JavaScript-only content).

L2 Structure fixes applied

  • Person schema sameAs updated · LinkedIn + Substack placeholder + GitHub.
  • LocalBusiness schema streetAddress removed · addressLocality «Barcelona» + addressCountry «ES» only.
  • Organization schema legalName field removed entirely (no entity declared per DEC-WEB-001).
  • All 13 routes pass Google Rich Results Test · 0 errors.
  • JSON-LD per page type per DEC-WEB-014 L2 spec — WebSite + Organization + Person + BreadcrumbList on Home · Article + HowTo + FAQPage on Methodology · Service + OfferCatalog (3 Offers) per tier page.

L3 Semantics fixes applied

  • Answer Capsules 300-400 words on every content-heavy page (Home · Methodology · 3 Service tier pages · Partners · 2 Case pages · 5 Field Notes · About).
  • Content sandwich pattern applied per Methodology M2 DEPT canon · median 377-word RAG chunks.
  • FLIP In-depth threshold (≥1500 words) met on all main pages.
  • Princeton GEO compliance per Methodology M4 — Statistics Addition (≥3 inline stats per page) · Quotation Addition (1-2 expert quotes where relevant) · Cite Sources (≥3 external authoritative links per page) · no Keyword Stuffing.
  • <dfn> definition discipline applied across all vAEO terms first-mention.

L4 Authority signal applied

  • Vlad's sameAs network: LinkedIn (canonical · linkedin.com/in/vladyslav-rovny) · Substack placeholder · GitHub.
  • External authoritative source links — Princeton GEO arXiv · Seer Interactive · Searchbloom · iPullRank · Profound · Gartner · Conductor · OpenAI · Anthropic across Methodology + Field Notes.
  • BBR case = real client proof of methodology.
  • Vlad's LinkedIn cross-linked back к vaeo.ai (bidirectional entity reinforcement).

Citation Stack composite self-score (post-fix · pre-launch):

LayerScore (0-100)BOX grade
L1 Access95 G
L2 Structure92 G
L3 Semantics87 G
L4 Authority71 S
Composite (Triangle Model weighted)86 G

«86 composite · Gold-grade Citation Stack health · self-applied · pre-launch. Same fix sequence we sell к clients. No special version for ourselves.»

AI crawlers actually visited.

Per Pre-Launch Gate C criterion 1 — AI crawler hit logs from Cloudflare must show key pages were visited within 14 days of staging deployment.

Cloudflare crawler hit log (staging deployment window · pre-launch):

CrawlerFirst hitPages visitedStatus
GPTBot (OpenAI)Day 2Home · Methodology · Services hub · BBR case
ClaudeBot (Anthropic)Day 4Home · Methodology · About · Field Notes
PerplexityBotDay 3Home · Methodology · Services hub · 3 tier pages
Google-ExtendedDay 5Home · Methodology · Services hub · About
BingbotDay 1Home · all 13 routes (full sitemap crawl)
OAI-SearchBotDay 6Home · Methodology · BBR case
Applebot-ExtendedDay 7Home only⚠️ partial
ChatGPT-User (user-triggered)observed Day 9Home · Services Diagnose

Gate C criterion 1 status: PASS — all 4 major AI bots (GPTBot · ClaudeBot · PerplexityBot · Google-Extended) visited key pages within 14 days. Bingbot bonus full-sitemap crawl. Applebot-Extended partial coverage flagged but не blocking.

Why crawler logs matter: crawler hit logs are the L1 Access discipline made observable. Without them · «we let AI engines crawl us» is just a claim. With them · we can prove robots.txt allowlist worked · llms.txt was discovered · sitemap was consumed.

«Eight crawlers · all visited · logs preserved. L1 Access discipline isn't a claim · it's a log entry.»

The probe — did AI engines cite us?

Per Pre-Launch Gate C criterion 2 — manual probe smoke test · vaeo.ai must be cited in ≥2 of 5 engines on ≥2 of 4 test queries within 14 days of staging indexing.

Probe methodology (T-2 weeks pre-launch · 14 days post staging deployment):

  • Engines probed: ChatGPT · Perplexity · Gemini · Google AI Overviews · Claude
  • Test queries (same 4 as baseline): «What is verifiable AEO?» · «Who does AI Visibility audits?» · «Best Citation Stack methodology» · «AEO agency partnership Barcelona»
  • Probe window: 14 days after staging deployment · before DNS cutover
  • Output capture: raw responses preserved as audit artifact

Self-AEO probe results (T-2w pre-launch):

Query ChatGPT Perplexity Gemini AIO Claude
«What is verifiable AEO?»✅ cited✅ citednot citednot cited✅ cited
«Who does AI Visibility audits?»✅ citednot citednot citednot cited✅ cited
«Best Citation Stack methodology»✅ cited✅ citednot citednot citednot cited
«AEO agency partnership Barcelona»not citednot citednot citednot citednot cited
Citations per engine3 of 42 of 40 of 40 of 42 of 4

Aggregated Share of Model (post-fix · pre-launch):

ChatGPT: 75% · Perplexity: 50% · Gemini: 0% · AIO: 0% · Claude: 50% · Cross-engine aggregated: 35%

Gate C criterion 2 status: PASS — vaeo.ai cited in 3 of 5 engines (ChatGPT · Perplexity · Claude) on 3 of 4 test queries. Threshold was ≥2 of 5 on ≥2 of 4 · met with margin.

Honest disclosures

  • Gemini and AIO cited zero times. Likely L4 Authority gap — Gemini relies heavily on Google Search index · vaeo.ai indexing window hadn't fully matured at probe time · AIO uses ranked organic results which weren't established yet. Continued monitoring post-launch.
  • «AEO agency partnership Barcelona» cited zero times across all engines. Likely L3 + L4 gap on geographic + service-type entity-graph clarity. Scheduled L4 reinforcement work post-launch.
  • Coverage = uneven. Methodology query («Best Citation Stack methodology») cited 2 of 5 · partnership-geographic query cited 0 of 5. Coverage shape is itself a finding.

«35% aggregated Share of Model · pre-launch · self-applied methodology. Gate C passed. Gemini · AIO · partnership-geographic query · still work to do. Honest about both.»

What's still broken · what comes next.

The dogfood case isn't a success story. It's a snapshot of a methodology applied honestly · including the parts that haven't compounded yet.

Currently underperforming

Gemini citation = 0 of 4 test queries. Gemini's reliance on Google Search index means newly-launched sites face longer indexing window before citation appears. Scheduled monitoring: weekly Gemini probe post-launch · expect citation appearance 30-60 days post-launch as Google indexing matures.

Google AI Overviews citation = 0 of 4. AIO uses ranked organic SERP positions as content base. vaeo.ai SERP rankings hadn't established at pre-launch probe. Scheduled monitoring: monthly AIO probe · expect citation 60-90 days post-launch as organic rankings build.

Geographic + service-type entity query («AEO agency partnership Barcelona») cited 0 of 5. L4 gap — vaeo.ai entity-graph hadn't established «Barcelona» + «partnership» + «AEO» cross-reinforcement at pre-launch. Scheduled L4 fix: Wikidata entity creation · industry directory placements · Substack publication launch · LinkedIn company page activation.

Wikidata entity for vAEO as Organization. Creation pending external review cadence (Wikidata community approval typically 2-6 weeks after submission). Scheduled post-launch.

Substack cross-link bidirectional. Placeholder pre-launch · will establish bidirectional reinforcement after Substack publication launches Q3 2026.

What success looks like in 90 days (post-launch target)

  • Aggregated Share of Model ≥ 50% across 5 engines on 4 test queries (vs 35% at launch)
  • Gemini citation ≥ 1 of 4 queries
  • AIO citation ≥ 1 of 4 queries
  • L4 composite Citation Stack score ≥ 85 (vs 71 at launch · L4 partial)

«35% at launch · 50% target at 90 days · Gemini + AIO measurement on probabilistic cadence. We name what doesn't work yet · we don't fake the chart.»

What we measure · monthly · forever.

Per Services-Dominate Dm2 Grow retainer cadence · self-applied: same monthly cadence vAEO sells к clients · applied к ourselves.

Monthly cadence (effective at vaeo.ai launch)

  • Week 1: Cross-engine probe (5 engines · 4 test queries + 10 expanded commercial queries) · Citation Stack composite health score · drift detection vs previous month.
  • Week 2: Proactive fix work on drift detected · schema corrections · sameAs maintenance · misrepresentation flags.
  • Week 3-4: Content sandwich rebuild on 1-2 priority pages (Field Notes new publications · methodology page refresh) · monthly authority brief drafted · monthly written report assembled.

Quarterly review (every Q)

  • 90-day Share of Model trend analysis
  • Citation Stack composite progression
  • L4 Authority signal maturity (sameAs · Wikidata · Substack · industry directories)
  • Next-quarter focus pivot decisions

Public update cadence on this page

  • Q3 2026: D7 + D8 sections updated with actual post-launch measurement results · new Share of Model numbers · success/limitation honest disclosure.
  • Q4 2026: 6-month engagement retrospective · what compounded · what didn't.
  • Annual: full case retrospective with year-over-year data.

«Monthly cadence · same as clients · same numbers reported back into the case page. Public · transparent · ongoing.»

The page that prevents the cobbler's-son problem.

The cobbler's son problem in AEO consulting is real and unsolved across the category. Vendors publish elaborate frameworks. Their own websites are invisible к AI engines. Schema validates with 12 errors. llms.txt absent. sameAs network = single LinkedIn link. Methodology not applied к the methodology owner.

The result: AEO industry credibility is fragile. Buyers can't distinguish vendors who do the work from vendors who write about doing the work.

This page is vAEO's load-bearing answer. Methodology applied к the methodology owner. Baseline · audit · fix sequence · crawler logs · probe results · limitations · continuous measurement plan · all preserved as audit artifact. Public. Verifiable. Updated monthly.

If vaeo.ai doesn't cite itself when AI engines are asked about AEO · we have no business charging for it. The dogfood case is that principle in load-bearing position.

«Cobbler's son has shoes. Methodology proven against the methodology owner. If we don't cite ourselves when AI is asked about AEO · we have no business charging for it.»