SEO + AEO + CRO Diagnosis

Revive & Align Pilates

www.reviveandalignpilates.com · 20 pages audited · 2026-08-12

63/100averaged over 20 measured pages

SEO84

AEO59

Brand entity40

CRO45

Overall · Reality check · critical

Your weakest front is Site audit (63/100, critical) — the fix order starts there, not where it's comfortable.

CRO: critical (45/100) — measured 2026-08-12.

Brand entity: mixed (40/100) — measured 2026-08-12.

Never measured: AI visibility (requires a tracking run) — you're flying blind there.

The numbers

pages audited

with launch-gating blockers

broken internal links

pages CRO-checked

40/100

entity signal score

unanswered buyer questions

Search & AI citability · Reality check · critical

1 of 20 audited pages carry a launch-gating blocker — those pages are dead weight until it's fixed.

There is almost nothing on this page. Search engines don't rank emptiness and AI can't quote what isn't written — right now this URL is a placeholder, not a page.

How to read the number: scores run 0–100 and a blocked page is hard-capped at 30/100 — the site sits at 63. This is not a polish problem, it's a locked gate: open it first or everything else is decoration.

These pages carry almost no entity markup — no schema-level identity an AI can verify you against (on-page entity markup 5/100). This is the page-level signal, not the brand-entity score reported separately.

Conversion · Reality check · critical

Your money pages are broken where it counts (site CRO 45/100 across 20 pages, high-intent weighted 3×).

5 high-intent page(s) carry a conversion-killing blocker — the funnel has holes exactly where the money flows.

Your worst money page is the lead_gen page at 45/100 (/) — that's where the traffic you paid for goes to die.

The most repeated leak: Clear value proposition — failing on 20 of 20 pages. One fix, applied everywhere.

Brand entity · Reality check · mixed

The entity layer is half-built (40/100) — recognizable, not yet authoritative.

Prioritized fix plan — one list, every lens

SEO, AEO, CRO, Entity and Content actions merged and ranked by severity. Checks marked heuristic in the underlying audits detect the attempt, not the quality.

Priority	Area	Fix	Pages	Source
critical	Critical	Resolve: Thin / low-information content Below the depth that ranks or gets cited, and adds no information gain — fails twice (LLMs self-generate generic content; humans ignore it).	1	Diamante Information Gain; Google HCU; KB §9
high	AEO	Add: Attributed expert quote ("Name, Title, says…") Quotations were the strongest GEO tactic for citation visibility (+41%, the only causal evidence). Caveat: 2,249 edit-tests (2026) put expert endorsements near-bottom for AI-RANKER relevance — treat quotes as a citation/trust lever, not a ranking lever.	20	Princeton/GT GEO, KDD 2024; DEJAN Jul 2026 (conflict — KB §32)
high	AEO	Add: Comparison table or structured list Comparison tables and listicles are the highest-cited formats in AI answers, and 'best-of' list framing was the single highest win-rate edit (38.4%) across 2,249 AI-ranking tests.	19	Wix AI Search Lab / Digital Bloom 2026; DEJAN Jul 2026; NP Digital Jul 2026
high	AEO	Add: Answer-first opening (self-contained answer in the first ~100 words) 44.2% of AI citations come from the first ~30% of a page; lead with a 40–60 word extractable answer.	1	Kevin Indig / Growth Memo, Mar 2026
high	SEO	Remove: Hidden text / cloaking signals Text hidden via display:none / font-size:0 / off-screen positioning is a classic cloaking/spam signal. (Low-confidence heuristic — verify it isn't legitimate accessible-hidden UI.)	2	Google spam policy (hidden text & cloaking)
high	CRO	Fix: Clear value proposition Failing on 20 of 20 audited page(s) (clarity).	20	CRO heuristics (NN/g · Baymard · CXL)
high	CRO	Fix: Social-proof numbers / logos Failing on 20 of 20 audited page(s) (trust).	20	CRO heuristics (NN/g · Baymard · CXL)
high	CRO	Fix: Security / credibility cues Failing on 20 of 20 audited page(s) (trust).	20	CRO heuristics (NN/g · Baymard · CXL)
high	CRO	Fix: Privacy + Terms links Failing on 20 of 20 audited page(s) (trust).	20	CRO heuristics (NN/g · Baymard · CXL)
high	CRO	Fix: CTA focus Failing on 19 of 20 audited page(s) (action).	19	CRO heuristics (NN/g · Baymard · CXL)
high	CRO	Fix: Guarantee / risk reversal Failing on 19 of 20 audited page(s) (trust).	19	CRO heuristics (NN/g · Baymard · CXL)
high	CRO	Fix: CTA early in page Failing on 19 of 20 audited page(s) (action).	19	CRO heuristics (NN/g · Baymard · CXL)
high	CRO	Fix: Scannable layout Failing on 10 of 20 audited page(s) (clarity).	10	CRO heuristics (NN/g · Baymard · CXL)
high	Entity	Create a Wikidata entry The entity backbone — free, and the highest-leverage AI-grounding signal.		KB §10
med	AEO	Add: Question-format H2/H3 headings Question headings match how people prompt AI and lift citation vs label-style headings.	3	Evertune 2026
med	SEO	Add: Descriptive <title> (~15–65 chars) The title is the #1 on-page relevance signal and the AI/SERP display line; lead it with the primary term.	5	KB §12 Pareto placement
med	AEO	Add: Visible, recent "last updated" date AI-cited pages skew ~26% fresher and ChatGPT is recency-biased — but the citation lift is query-intent-dependent: strong for time-sensitive / commercial topics, weaker for evergreen / definitional content. Refresh ~every 2 months with a visible date where recency matters; don't force a fake update on genuine reference pages.	2	Ahrefs 2025 / Seer 2025; SIGIR 2026 + Martinez survey — conditional (KB §34)
med	SEO	Fix: duplicate H1 headings across pages Duplicate H1s blur each page's primary topic for parsers and readers.	20	KB §12 Pareto placement
med	AEO	Add: Named author (byline / author schema) E-E-A-T: an attributable human author (visible byline or Person/author schema) is a core trust signal for both Google and AI grounding; anonymous content reads as content-farm output.	20	Google E-E-A-T guidelines; KB §6/§20 (named-author content earns AI citations)
med	AEO	Add: Sections open with a self-contained answer Each section should open with a sentence that answers its heading and stands alone (no unresolved 'this/that/above'). Answer-first passages carrying their own headings are the one causally-proven on-page sub-tactic — this is NOT content-chunking.	1	KDD 2024 / KB §29; Zyppy Self-Contained Passages 8.0, May 2026
med	SEO	Refresh 18 stale page(s) (>2 months old) AI-cited pages skew ~26% fresher and ChatGPT is recency-biased; refresh content and bump the visible date roughly every 2 months.	18	Ahrefs 2025 / Seer 2025; KB §5
med	AEO	Add: Leads with original / first-party data (Information Gain) Original data is the moat — necessary for citation, not sufficient alone. The winning FORM is a comparison benchmark answering 'which is best': primary research is only ~2.7% of AI-cited pages but earns 3.3× the citation density, and nearly all of it is benchmarks.	20	Diamante 'Get Found' ~Dec 2025 & KB §9; Indig/Growth Memo Jun 29 + Jul 6 2026
med	Entity	Create/link profiles: LinkedIn, Wikidata, Crunchbase, YouTube, Google Business Consistent third-party profiles anchor the entity and are where AI looks for corroboration.		KB §10/§14
low	SEO	Add: Meta description present (~70–160 chars) Controls the SERP/AI preview snippet; a missing one cedes the framing to the engine.	10	KB §12; Cyrus Shepard preview-control
low	Entity	Add: Organization schema with sameAs (entity signals) Entity verification gates whether citations 'stick' (and drives rich results). Keep schema — just not as a citation trick.	19	KB Entity Layer §10

E-E-A-T pillars

Experience · Expertise · Authoritativeness · Trust — Google's Quality-Rater vocabulary. There is no single E-E-A-T score (Google says so): four separate readouts. Experience & Expertise are detected-signal checklists — we see the signal, not the substance.

Experience — 1/4 signals detected — checklist, not a score — First-hand, real-world involvement with the topic — the page shows the author actually did, used, or tested the thing, with original observations, results, or 'we found…' data rather than a rewrite of what's already online.

Google's Quality Rater Guidelines added the extra 'E' for Experience: for many queries the most helpful result comes from someone who has genuinely lived it, and AI answers increasingly favour pages that read as first-hand.

We detect the SIGNAL, not the substance — we can see first-person, tested-it phrasing and case-study structure, but we cannot verify the author truly has the experience. Read these as attempts present or absent, not proof.

✓ Visible, recent "last updated" date heuristic weight 2

Found: 18 of 20 content pages

Fix: Add a visible, machine-parseable last-updated date (and article:modified_time); keep it genuinely current.

Why: AI-cited pages skew ~26% fresher and ChatGPT is recency-biased — but the citation lift is query-intent-dependent: strong for time-sensitive / commercial topics, weaker for evergreen / definitional content. Refresh ~every 2 months with a visible date where recency matters; don't force a fake update on genuine reference pages.

Ahrefs 2025 / Seer 2025; SIGIR 2026 + Martinez survey — conditional (KB §34)

✕ Leads with original / first-party data (Information Gain) heuristic weight 2

Found: 0 of 20 content pages

Fix: Lead with original / first-party data or a benchmark others don't have — the moat for citation.

Why: Original data is the moat — necessary for citation, not sufficient alone. The winning FORM is a comparison benchmark answering 'which is best': primary research is only ~2.7% of AI-cited pages but earns 3.3× the citation density, and nearly all of it is benchmarks.

Diamante 'Get Found' ~Dec 2025 & KB §9; Indig/Growth Memo Jun 29 + Jul 6 2026

✕ First-hand experience heuristic weight 2

Found: 0 of 20 content pages

Fix: Add first-person, tested-it detail — 'in our experience…', 'when we tested…', original observations and numbers — instead of a rewrite of what's already online.

Why: Pages that show the author actually did / used / tested the thing read as first-hand — the lever Google's Experience 'E' and AI answers reward over rewrites.

Google Quality Rater Guidelines (Experience); AEO KB §6

✕ Case study / results heuristic weight 2

Found: 2 of 20 content pages

Fix: Add a case study or results section — 'we helped X achieve Y', a before-and-after, or a metric-backed outcome.

Why: A case-study or before/after results structure is concrete proof of real experience and a strong citation magnet.

AEO KB §6

Expertise — 0/4 signals detected — checklist, not a score — Whether the page is attributed to a credentialed, identifiable person — a visible byline linked to a real author bio, a Person schema node with a job title and profiles, an attributed expert quote, or a 'reviewed by' line.

AI answers and Google both weigh who stands behind the content; anonymous content competes worse, and for many topics an unnamed author is a citation dead-end.

Signal, not substance — we can detect a byline, a bio link, and a reviewer line, but we cannot verify the person is genuinely an expert. These are detected attempts, not a competence score.

✕ Attributed expert quote ("Name, Title, says…") heuristic weight 3

Found: 0 of 20 content pages

Fix: Add an attributed quote from a named expert ("according to — Jane Smith, …").

Why: Quotations were the strongest GEO tactic for citation visibility (+41%, the only causal evidence). Caveat: 2,249 edit-tests (2026) put expert endorsements near-bottom for AI-RANKER relevance — treat quotes as a citation/trust lever, not a ranking lever.

Princeton/GT GEO, KDD 2024; DEJAN Jul 2026 (conflict — KB §32)

✕ Named author (byline / author schema) heuristic weight 2

Found: 0 of 20 content pages

Fix: Add a visible byline linked to an author bio page + a Person author node.

{"@type":"Person","name":"…","url":"https://…/about/jane","sameAs":["https://linkedin.com/in/…"]}

Why: E-E-A-T: an attributable human author (visible byline or Person/author schema) is a core trust signal for both Google and AI grounding; anonymous content reads as content-farm output.

Google E-E-A-T guidelines; KB §6/§20 (named-author content earns AI citations)

✕ Author bio + Person schema measured weight 2

Found: 0 of 20 content pages

Fix: Add a Person author node with a bio-page url and at least one of jobTitle / sameAs / knowsAbout.

{"@type":"Person","name":"…","url":"https://…/about/jane","jobTitle":"…","sameAs":["https://linkedin.com/in/…"]}

Why: A named author linked to a real bio via a Person schema node (job title / profiles) proves expertise a machine can resolve — anonymous content is a citation dead-end.

Google Quality Rater Guidelines (Expertise); AEO KB §6

✕ Reviewed-by line heuristic weight 1

Found: 0 of 20 content pages

Fix: Add a visible 'Reviewed by <named expert>' or 'Fact-checked by …' line where accuracy matters (a bonus on non-YMYL pages).

Why: A 'reviewed by / fact-checked by <expert>' line signals editorial oversight — especially important for YMYL (your-money-your-life) topics.

Google Quality Rater Guidelines (Expertise/Trust)

Authoritativeness — 3/7 present · 33% coverage (3 unmeasured) — Whether the wider web treats the brand as a recognised authority: a resolvable Organization entity (Org schema, Wikidata/Wikipedia, consistent profiles), a substantive About page, authoritative outbound citations, a real internal-link structure, and off-site backlinks / referring domains.

Authoritativeness is reputation earned off the page — the hardest E-E-A-T pillar to fake and the one AI answers lean on to decide whom to name. An unrecognised brand gets paraphrased away.

✓ In-body internal links (topical depth) measured weight 1

Found: 20 of 20 pages

Fix: Add 2–4 in-body links to related pages on your site (hub↔spoke) using descriptive anchor text.

Why: Internal linking signals topical authority (hub↔spoke) and is the mechanism that compounds cluster citations.

KB §13 topical authority

✓ 2–4 outbound links to authoritative sources measured weight 1

Found: 20 of 20 pages

Fix: Add 2–4 outbound links to authoritative sources (research/data, not competitors).

Why: Outbound links to real research/data correlate with stronger, more trustworthy pages.

E. Sturm experiment 2026; KB §12

✕ Organization schema with sameAs (entity signals) measured weight 1

Found: 1 of 20 pages

Fix: Add an Organization node with sameAs to your real profiles. For full entity depth (Wikidata/Wikipedia/ownership) run the Brand-entity check.

{"@type":"Organization","name":"…","url":"https://…","sameAs":["https://…"]}

Why: Entity verification gates whether citations 'stick' (and drives rich results). Keep schema — just not as a citation trick.

KB Entity Layer §10

✕ Substantive About page measured weight 2

Found: no About / company page found in the crawl

Fix: Publish a substantive About page (team, story, mission, credentials) and link it from the main nav.

Why: A real About/company page is a core authority signal — it lets people and machines establish who is behind the site.

Google Quality Rater Guidelines (Authoritativeness)

✕ Wikidata entity (owned) measured weight 2

Found: no Wikidata entry found

Fix: Create a Wikidata item for your organisation with an official-website (P856) value pointing to your domain. If an entry exists but its P856 doesn't match yours, it is a DIFFERENT entity — create your own; don't claim theirs.

Why: A Wikidata item that YOU own (its official-website claim P856 points to your domain) is the strongest machine-readable proof your brand is a real, notable entity — AI answers resolve brands to Wikidata / the Knowledge Graph.

AEO KB (entity / Wikidata); Google Knowledge Graph

✕ Wikipedia article measured weight 1

Found: no Wikipedia entry found

Fix: Earn a Wikipedia article through genuine notability and independent third-party coverage — do not create a promotional stub (it will be removed).

Why: A Wikipedia article is a top notability + authority signal and a frequent source AI assistants cite.

AEO KB (entity); Google Knowledge Graph

✓ Linked profiles (sameAs) measured weight 1

Found: 2 live profile(s) of 3 linked

Fix: Add an Organization sameAs array on your homepage linking to your real, live profiles (LinkedIn, Crunchbase, X, GitHub, etc.).

Why: Consistent, live links from your homepage to your real profiles (LinkedIn, Crunchbase, GitHub, socials) via Organization sameAs help machines connect and corroborate your entity.

AEO KB (entity / sameAs)

? Domain Rating (backlink authority) measured weight 2

Found: not measured — no AHREFS_API_KEY configured

Fix: Earn links from reputable, relevant sites (original data, useful tools, genuine PR) — not link schemes. DR climbs with quality referring domains over time.

Why: Domain Rating summarises the strength of your backlink profile (0-100). It's off-page reputation earned across the web — the hardest E-E-A-T signal to fake and a strong correlate of whom AI answers trust.

Ahrefs REST API v3 (site-explorer/domain-rating)

? Referring domains measured weight 1

Found: not measured — no AHREFS_API_KEY configured

Fix: Grow the number of DISTINCT reputable domains that link to you (guest expertise, data others cite, integrations, listings) rather than many links from one site.

Why: The number of unique domains linking to you is a more robust authority signal than raw backlink count — breadth of endorsement.

Ahrefs REST API v3 (site-explorer/backlinks-stats.live_refdomains)

? Backlinks measured weight 1

Found: not measured — no AHREFS_API_KEY configured

Fix: Attract more links to your best pages (linkable assets, original research, tools) — quality and referring-domain breadth matter more than raw count.

Why: Total live backlinks is a raw volume signal of off-page endorsement; most useful read alongside referring domains.

Ahrefs REST API v3 (site-explorer/backlinks-stats.live)

Trust — 3/8 present · 55% coverage (1 unmeasured) — The signals that tell a visitor (and a machine) the site is safe and legitimate: HTTPS, real reviews / testimonials and ratings, guarantees or warranties, security and accreditation cues, a name / address / phone, a reachable contact path, and privacy + terms pages.

Google calls Trust the centre of E-E-A-T — the other three pillars only matter if the page is trustworthy. It is also the pillar that most directly drives conversion.

✓ Machine-readable pricing heuristic weight 2

Found: 19 of 19 pages

Fix: On commercial pages, show real prices in the page copy AND mark them up with Offer / PriceSpecification schema so AI can extract them — don't hide pricing behind 'contact sales' if you want to be cited for commercial queries.

{"@type":"Offer","price":"29.00","priceCurrency":"USD","priceSpecification":{"@type":"UnitPriceSpecification","unitText":"MONTH"}}

Why: AI answer engines cite pages with explicit, extractable pricing far more often — the largest controlled study to date (SIGIR 2026 'What Gets Cited', 252k trials) and Indig's Jul-2026 agent-friction experiment both find pricing a top citation lever; when your pricing isn't on-site, agents answer from directories (G2/Capterra) instead of you. Transparent pricing is also a buyer-trust signal. (Only checked on commercial pages — n/a on articles/about/legal.)

SIGIR 2026 (What Gets Cited); Growth Memo (Indig, Jul 13 2026); AEO KB §33–§34 (lift strongest on commercial / time-sensitive intent)

✕ Name / address / phone heuristic weight 1

Found: 0 of 20 pages

Fix: Show a consistent business name, postal address and phone number (ideally in the footer plus a contact page).

Why: A consistent business name, postal address and phone is a core local-trust and entity signal for a real organisation.

AEO KB (entity / NAP consistency)

✓ HTTPS measured weight 2

Found: all 20 measured page(s) served over HTTPS

Fix: Serve every page over HTTPS with a valid certificate and 301-redirect http→https.

Why: HTTPS is a baseline trust and ranking signal; a page served over http is labelled 'not secure' to users.

Google (HTTPS as a ranking signal)

? Third-party review-site presence heuristic weight 1

Found: could not confirm review-site presence — checked G2/Trustpilot/Clutch/Capterra heuristically

Fix: Claim and populate profiles on the review platforms relevant to your category (G2/Capterra for software, Clutch for agencies, Trustpilot broadly) and earn genuine reviews.

Why: A presence on independent review platforms (G2, Trustpilot, Clutch, Capterra) is off-site social proof that both buyers and AI assistants use to gauge legitimacy.

AEO KB (off-page trust); public review-site profile checks (best-effort)

✓ Testimonials / reviews heuristic weight 2

Found: 20 of 20 pages

Why: Social proof reduces risk

Baymard

✕ Social-proof numbers / logos heuristic weight 1

Found: 0 of 20 pages

Fix: Social-proof numbers / logos

Why: Specific proof beats claims

CXL

✕ Guarantee / risk reversal heuristic weight 1

Found: 1 of 20 pages

Fix: Guarantee / risk reversal

Why: Removes purchase risk

CXL

✕ Security / credibility cues heuristic weight 1

Found: 0 of 20 pages

Fix: Security / credibility cues

Why: Legitimacy signals raise trust

Baymard

✕ Privacy + Terms links measured weight 1

Found: 0 of 20 pages

Fix: Privacy + Terms links

Why: Legitimacy + compliance

NN/g

Off-page opportunities

How to earn authority off-site — ranked quick-wins first. For AI visibility, brand mentions & earned media beat raw backlinks; these are link-earning moves (editorial only), not link-building.

○ Create a Wikidata item medium effort · high impact

Create a Wikidata item for your org with official website (P856) → your domain.

An owned, machine-readable Wikidata entity is the strongest entity proof; AI resolves brands to Wikidata / the Knowledge Graph. — AEO KB §10 / §33 (entity)

○ Earn a Wikipedia article high effort · high impact

Earn a Wikipedia article through genuine notability + independent third-party coverage (a promo stub will be removed).

A Wikipedia article is a top notability signal and a frequent AI citation source. — AEO KB §10 (entity)

? Check your G2 profile low effort · high impact

Couldn't confirm your G2 presence (probe blocked or no match) — verify manually, and claim + fully populate a profile if you don't have one.

Review-site profiles are borrowed authority AI already cites (G2 is ~#4-cited overall) — the fastest off-page win when you have no existing links. — AEO KB §14 (borrowed authority)

? Check your Trustpilot profile low effort · high impact

Couldn't confirm your Trustpilot presence (probe blocked or no match) — verify manually, and claim + fully populate a profile if you don't have one.

Review-site profiles are borrowed authority AI already cites (G2 is ~#4-cited overall) — the fastest off-page win when you have no existing links. — AEO KB §14 (borrowed authority)

? Check your Clutch profile low effort · high impact

Couldn't confirm your Clutch presence (probe blocked or no match) — verify manually, and claim + fully populate a profile if you don't have one.

Review-site profiles are borrowed authority AI already cites (G2 is ~#4-cited overall) — the fastest off-page win when you have no existing links. — AEO KB §14 (borrowed authority)

? Check your Capterra profile low effort · high impact

Couldn't confirm your Capterra presence (probe blocked or no match) — verify manually, and claim + fully populate a profile if you don't have one.

Review-site profiles are borrowed authority AI already cites (G2 is ~#4-cited overall) — the fastest off-page win when you have no existing links. — AEO KB §14 (borrowed authority)

? Backlink authority — not measured high effort · high impact

Set AHREFS_API_KEY to measure Domain Rating / referring domains, then publish an original-data asset to close the gap.

Backlink authority from real editorial still feeds ranking + E-E-A-T; measure it to prioritise a linkable-data asset. — AEO KB §6.3 / §26

? Reclaim unlinked brand mentions low effort · medium impact

Find pages that mention your brand without linking, and request the link — the cheapest link there is. (Automating this needs a search key.)

Unlinked mentions are earned coverage that already exists; converting them to links is pure upside. — AEO KB §14

? Find competitor referring-domain gaps medium effort · medium impact

Not measured — no competitors provided (pass --vs <url> or configure competitors). Set AHREFS_API_KEY + pass competitors (--vs <url>) to find domains linking to rivals but not you.

Referring domains linking to rivals but not you are the warmest link targets. — AEO KB §26

✓ Linked profiles (sameAs) low effort · medium impact

Off-page & entity checklist (not crawl-detectable — verify manually)

Wikidata entry exists (the entity backbone; free, highest-leverage — do before content). [KB §10 Entity Layer]
Organization schema `sameAs` links to LinkedIn, Wikidata, and your review/directory profiles. [KB §10]
NAP (name / contact) is identical across the site, Google Business Profile, LinkedIn, directories. [KB §10]
A directory/review profile is live with ≥1 review (Clutch for B2B, G2 for software, Google). [KB §10 / §14]
Third-party brand mentions + earned media in motion (press, expert quotes) — the #1 AEO lever. [Ahrefs 0.664; Muck Rack 84%]
A YouTube presence for your buyer topics (highest single citation correlation ~0.737). [Ahrefs Dec 2025]
Publish named-author LinkedIn content — LinkedIn is now the #2 most-cited AI domain (~11%, rising while Reddit falls); post as a person, not the company page. Wikipedia / YouTube / Medium round out the top sources; Reddit is #1 but declining — authentic participation only, never seeding. (Your `track` run's cited-domain breakdown personalizes this to your own buyer questions.) [Semrush Most-Cited Domains; Otterly, 2026 — shifting, re-check]

Scope & methodology

What this diagnosis did and didn't measure

Measured: on-page SEO/AEO signals across the crawled pages (blockers cap scores), conversion heuristics, entity presence (Wikidata/Wikipedia/schema/profiles). Not measured in this run: live AI-citation share (requires a tracking run across ChatGPT/Perplexity/Gemini/AI Overviews), real Core Web Vitals (optional PSI run), and backlink/authority data. Every check cites a source; scoring model per the SEO/AEO knowledge base v2.6.

• Schema is for rich results + entity clarity, NOT an AI-citation lever (no measurable lift in a 1,885-page test). Keep it; don't over-invest. [Ahrefs, May 2026]
• Don't chase a fixed 'AI rank position' — answers are inconsistent run-to-run. Track inclusion / share-of-voice instead. [Fishkin/Gumshoe, Jan 2026]
• The biggest AEO lever is OFF-page: third-party brand mentions + earned media out-predict backlinks for AI visibility. Pair on-page work with PR/YouTube/reviews. [Ahrefs 75k-brand study; Muck Rack 84%]
• Lead with information gain — and shape it as the category BENCHMARK ('which is best'). Primary research is ~2.7% of AI-cited pages but takes 3.3× the citations; verticals with no benchmark produce zero cited research. [Diamante; Indig/Growth Memo Jul 6 2026; KB §9/§32]

Prepared by AEO Studio · 2026-08-12