Start by auditing whether your brand appears in relevant AI-generated answers, then repair the pages that lack clear entities, extractable answers, citations, and topical depth.
GEO improvement depends on making authoritative content easy for generative systems to retrieve, summarize, and attribute. The work splits into two failure modes: retrieval gaps (the engine never fetched the right page) and interpretation gaps (it fetched and misread). They need different fixes.
,
fix: { tag: 'how SurfaceGX repairs this', text: <>The Surface Audit module benchmarks visibility across AI engines and produces a prioritized repair plan., target: 'modules' },
},
{
id: 'improve-aeo-citations',
cat: 'aeo', catLabel: 'AEO',
q: "How can my company improve AEO so answer engines cite our content?",
text: "Build answer-first content that directly resolves specific user questions, supported by structured data, expert validation, and internal links to deeper resources. AEO works best when each page provides concise, factual, and reusable answers: a paragraph an engine can lift verbatim, backed by evidence the engine can verify.",
a: <>
Build answer-first content that directly resolves specific user questions, supported by structured data, expert validation, and internal links to deeper resources.
AEO works best when each page provides concise, factual, and reusable answers: a paragraph an engine can lift verbatim, backed by evidence the engine can verify.
,
fix: { tag: 'how SurfaceGX repairs this', text: <>The Content Brief Empower module ships writer-ready briefs that target each engine's interpretation failure with cited evidence., target: 'modules' },
},
{
id: 'check-ai-readiness',
cat: 'ready', catLabel: 'Readiness',
q: "How do I know if my site is AI-ready for chatbots and generative search?",
text: "Your site is AI-ready when key pages are crawlable, renderable, indexable, semantically structured, and supported by accurate schema and entity signals. Test whether important content can be extracted from the HTML without relying on blocked scripts or inaccessible assets. If a curl request doesn't return your value prop, neither does an AI engine.",
a: <>
Your site is AI-ready when key pages are crawlable, renderable, indexable, semantically structured, and supported by accurate schema and entity signals.
Test whether important content can be extracted from the HTML without relying on blocked scripts or inaccessible assets. If a curl request doesn't return your value prop, neither does an AI engine.
,
fix: { tag: 'how SurfaceGX repairs this', text: <>The Surface Audit runs all five readiness checks and ships fix cards for any failures., target: 'modules' },
},
{
id: 'improve-ai-readability',
cat: 'read', catLabel: 'Readability',
q: "What changes improve my site's AI readability for LLMs and search crawlers?",
text: "Use descriptive headings, concise definitions, schema markup, tables, FAQs, author information, and consistent terminology. Remove ambiguity by clearly connecting claims, entities, products, services, and supporting evidence. Engines reward pages that read like a structured argument, not a brochure.",
a: <>
Use descriptive headings, concise definitions, schema markup, tables, FAQs, author information, and consistent terminology.
Remove ambiguity by clearly connecting claims, entities, products, services, and supporting evidence. Engines reward pages that read like a structured argument, not a brochure.
,
fix: { tag: 'how SurfaceGX repairs this', text: <>The Engineering Handoff module ships brand manifests, llms.txt, and schema patches as deployable PRs., target: 'modules' },
},
{
id: 'javascript-pdf-gated',
cat: 'tech', catLabel: 'Technical',
q: "Can AI chatbots read my site if it uses JavaScript, PDFs, or gated content?",
text: "AI systems may miss content that is only available through heavy JavaScript, PDFs, forms, logins, or blocked resources. Publish critical information in accessible HTML and ensure crawlers are not restricted from essential page elements. SSR or hydration-safe HTML is a baseline; PDFs should be paired with HTML equivalents; gated content needs a public summary engines can ingest.",
a: <>
AI systems may miss content that is only available through heavy JavaScript, PDFs, forms, logins, or blocked resources.
Publish critical information in accessible HTML and ensure crawlers are not restricted from essential page elements. SSR or hydration-safe HTML is a baseline; PDFs should be paired with HTML equivalents; gated content needs a public summary engines can ingest.
,
fix: { tag: 'how SurfaceGX repairs this', text: <>The Engine Diagnosis module pinpoints which assets your visibility-critical pages actually depend on, and which engines can't reach them., target: 'modules' },
},
{
id: 'improve-geo-generative-search',
cat: 'geo', catLabel: 'GEO',
q: "How do I improve my GEO and visibility in generative search?",
text: "Focus on source attribution: cite authoritative data and use industry-specific terminology that aligns with the training and retrieval surfaces of major LLMs. Generative engines prefer pages with information density over keyword density. A page with three crisp definitions, two charts, and a citation outranks a page with the same word stuffed twenty times.",
a: <>
Focus on source attribution: cite authoritative data and use industry-specific terminology that aligns with the training and retrieval surfaces of major LLMs.
Generative engines prefer pages with information density over keyword density. A page with three crisp definitions, two charts, and a citation outranks a page with the same word stuffed twenty times.
,
fix: { tag: 'how SurfaceGX repairs this', text: <>The Surface Audit scores source attribution against authoritative-citation weights tuned per engine., target: 'modules' },
},
{
id: 'improve-aeo-voice-chat',
cat: 'aeo', catLabel: 'AEO',
q: "How do I improve my company's AEO for voice and chat results?",
text: "Optimize for question-based queries and supply direct, data-backed answers within your HTML to become the preferred source for LLM-generated responses. Voice and chat surfaces strongly prefer 40–60 word answers with a single supporting clause. Structure your content so the first paragraph of every section is the extractable answer.",
a: <>
Optimize for question-based queries and supply direct, data-backed answers within your HTML to become the preferred source for LLM-generated responses.
Voice and chat surfaces strongly prefer 40–60 word answers with a single supporting clause. Structure your content so the first paragraph of every section is the extractable answer.
,
fix: { tag: 'how SurfaceGX repairs this', text: <>Content briefs include voice-and-chat optimized snippets per page, cited to the engine that demanded them., target: 'modules' },
},
{
id: 'ai-ready-next-gen-search',
cat: 'ready', catLabel: 'Readiness',
q: "How do I know if my site is AI-ready for the next generation of search?",
text: "Conduct a semantic audit to ensure your content is entity-mapped and that your technical stack permits low-latency access for AI crawlers like GPTBot or OAI-SearchBot. The next generation of engines query crawl-on-demand. If your TTFB is high or you rate-limit AI user agents, you're invisible, even with perfect schema.",
a: <>
Conduct a semantic audit to ensure your content is entity-mapped and that your technical stack permits low-latency access for AI crawlers like GPTBot or OAI-SearchBot.
The next generation of engines query crawl-on-demand. If your TTFB is high or you rate-limit AI user agents, you're invisible, even with perfect schema.
AI engines are misreading your brand.
Repair the input layer they read.
SurfaceGX is AI visibility repair infrastructure. We verify the technical input layer
AI engines read (fetchability, schema, headings, canonicals, robots, sitemap, llms.txt)
and use LLMs only where model interpretation is required.
01Diagnose where the breakdown is happening across owned pages, third-party sources, and retrieval surfaces.
02Separate verified technical evidence from LLM interpretation. Your team knows which fixes are infrastructure and which require judgment.
03Ship repairs as Fix Cards, manifests, dev handoffs, and GitHub PR workflows, not screenshots.
When someone asks an AI assistant about your category, your brand may be invisible,
inaccurately described, or missing from the answer entirely. Traditional SEO tools and
AEO monitoring dashboards were not built to repair this. Tracking the output
does not fix the input layer.
setActive('concepts')}>
deeper read
Want the precise definitions of GEO, AEO, retrieval surface and how llms.txt
changes ingestion? They live in Key Concepts.
Key Concepts →
);
}
/* ── Section: Modules ─────────────────────────────────────────────────── */
function SectionModules({ setActive }) {
const modules = [
{ num: '01', tag: 'Audit', tagCls: 'cyan', accent: 'cyan',
title: 'Surface Audit',
desc: 'Crawl owned pages, retrieval surfaces, and third-party mentions. Score fetchability, schema coverage, canonicals, robots/sitemap, llms.txt, and structural clarity.',
output: 'Outputs: full evidence ledger + score matrix' },
{ num: '02', tag: 'Diagnose', tagCls: 'violet', accent: 'violet',
title: 'Engine Diagnosis',
desc: 'Ask the answer engines what they actually retrieved. Differentiate retrieval failures from interpretation failures using model-specific probes.',
output: 'Outputs: model-vs-truth deltas, source heatmap' },
{ num: '03', tag: 'Empower', tagCls: 'green', accent: 'green',
title: 'Content Brief Empower',
desc: 'For interpretation gaps, generate prescriptive content briefs your writers can act on. Each brief cites the engine that misread you and why.',
output: 'Outputs: writer-ready briefs, evidence pins' },
{ num: '04', tag: 'Repair', tagCls: 'amber', accent: 'amber',
title: 'Engineering Handoff',
desc: "For infrastructure gaps, ship Fix Cards, generated manifests, and GitHub PRs against your repos. Your engineers merge — they don't translate.",
output: 'Outputs: Fix Cards, manifests, PR diffs' },
];
return (
Platform · Modules
Four modules, one repair loop
Each module produces durable artifacts your team can act on. No dashboard-only insights.
Every score has evidence; every recommendation has a deployable output.
{modules.map(m => (
{m.tag}M.{m.num}
{m.title}
{m.desc}
→ {m.output}
))}
setActive('concepts')}>
definitionsGEO, AEO, retrieval surfaces, llms.txt — the public concepts each module
operates against live in Key Concepts.
Key concepts →
);
}
/* ── Section: Outputs ─────────────────────────────────────────────────── */
function SectionOutputs({ setActive }) {
const outputs = [
{ num: 'O.01', name: 'Fix Card', desc: "A single repair: what's broken, evidence, recommended diff, owner, effort. Filterable, exportable, mergeable.", tag: 'JSON · MD · PR' },
{ num: 'O.02', name: 'Brand Manifest', desc: 'Canonical description AI engines should read. Versioned, signed, served at /brand-manifest.json.', tag: 'JSON-LD' },
{ num: 'O.03', name: 'llms.txt', desc: 'Generated and validated llms.txt + llms-full.txt for your domain, scoped to what engines should actually ingest.', tag: 'TEXT' },
{ num: 'O.04', name: 'GitHub PR', desc: 'Repair patches opened directly against your repos with full evidence in the PR body. CI-friendly.', tag: 'GIT' },
{ num: 'O.05', name: 'Content Brief', desc: "Writer-ready brief targeting one engine's interpretation failure. Includes citations, voice notes, and structure.", tag: 'MD · DOCX' },
{ num: 'O.06', name: 'Evidence Ledger',desc: 'Append-only log of every check, score, and source. Court-of-record for the program — auditable by anyone.', tag: 'JSONL' },
];
return (
Platform · Outputs
What it ships, not what it shows
Dashboards report. SurfaceGX repairs. Every module produces deployable artifacts: Fix Cards,
manifests, and PRs your team can review, version, and merge.
{outputs.map(o => (
{o.num}{o.name}{o.desc}{o.tag}
))}
);
}
/* ── Section: Storyline ───────────────────────────────────────────────── */
function SectionStoryline({ setActive }) {
const steps = [
{ num: 'STEP 01', h: 'Evidence first', body: 'Crawl the surfaces engines actually retrieve. Capture fetchability, schema, robots, sitemap, llms.txt as ground truth.', foot: 'no models yet' },
{ num: 'STEP 02', h: 'Ask the engines', body: 'Query GPT, Claude, Gemini and others with brand-relevant prompts. Compare the answer to the evidence: what did they get wrong?', foot: 'model-as-witness' },
{ num: 'STEP 03', h: 'Classify the failure', body: "Each gap is either an infrastructure failure (page wasn't fetched cleanly) or an interpretation failure (page was read but misjudged).", foot: 'infra vs interp' },
{ num: 'STEP 04', h: 'Generate the repair', body: 'Infra failures become Fix Cards + PRs. Interp failures become content briefs. Each repair cites the engine that flagged it.', foot: 'artifacts, not advice' },
{ num: 'STEP 05', h: 'Re-scan and verify', body: 'After deploy, re-run the same probes. Score the delta. Close the Fix Card with evidence. Repeat for the next quarter.', foot: 'measurable loop' },
];
return (
Strategy · How it works
A five-step repair loop
SurfaceGX is opinionated about sequence. Evidence before model interpretation. Classification before
recommendation. Repair before re-measurement. The loop is the product.
{steps.map(s => (
{s.num}
{s.h}
{s.body}
— {s.foot}
))}
Why this order matters
01
Models can't witness their own ingestion.
If you ask a model what it knows, you get its memory, not what's currently fetchable. Crawl first, ask second.
02
Interpretation gaps and infra gaps look identical from a dashboard.
Both show "the AI got it wrong." Repairing the wrong layer wastes a quarter. Classification is the unlock.
03
A repair only counts when it merges.
Insights die in dashboards. PRs ship. The platform optimizes for deployable artifacts, not screenshot-able charts.
SurfaceGX helps teams move from measurement to repair. The platform diagnoses the underlying technical, content, and authority signals that determine whether AI engines can discover, parse, trust, and cite a brand, then turns those findings into deployable fixes for marketing, SEO, and engineering teams.
SEO PLATFORMS
AEO MONITORS
SurfaceGX
{rows.map((r, i) => (
{r.label}
{r.seo}
{r.aeo}
{r.sg}
))}
setActive('overview')}>
comparison
A vendor-by-vendor breakdown (Meltwater GenAI Lens, Profound, SE Ranking, AthenaHQ)
lives in the platform overview.
Platform overview →
);
}
/* ── Section: Teams ───────────────────────────────────────────────────── */
function SectionTeams({ setActive }) {
const aud = [
{ role: 'CMO / Marketing leader', cls: 'tag-cyan', accent: 'cyan',
title: 'Move beyond AI visibility reporting.',
pain: "You can see you're losing mentions. You can't see where the input layer broke. Quarterly review needs a repair plan, not another dashboard.",
fix: 'Quarterly repair plan → Fix Cards with owners, effort, and merge status.' },
{ role: 'Content lead', cls: 'tag-green', accent: 'green',
title: 'Stop guessing what to write next.',
pain: "AEO tools say \"you're not mentioned in answers.\" They don't tell your writers what page to rebuild or what angle to take.",
fix: 'Engine-specific briefs → cited interpretation gaps, voice notes, structural skeleton.' },
{ role: 'SEO / Digital team', cls: 'tag-violet', accent: 'violet',
title: 'Extend ranking work into the input layer.',
pain: "You've already optimized crawl. Now half your traffic is mediated by AI. The work isn't the same and the tools don't cover it.",
fix: 'Manifest + llms.txt + schema gaps shipped as PRs your dev already trusts.' },
{ role: 'Agency / Multi-brand', cls: 'tag-amber', accent: 'amber',
title: 'Run GEO/AEO programs at scale.',
pain: "Per-client dashboards don't compare. Your team can't demonstrate program ROI across a portfolio without a unified output format.",
fix: 'Workspace-per-client + cross-portfolio repair ledger + white-label PRs.' },
];
return (
Go-to-market · Teams
Built for the teams who own the fix
SurfaceGX is one platform with four working surfaces. Each role gets artifacts shaped for
their workflow, not a dashboard everyone has to translate from.
{aud.map(a => (
{a.role}
{a.title}
{a.pain}
What they get → {a.fix}
))}
);
}
/* ── Section: Services / pricing ──────────────────────────────────────── */
function SectionServices({ setActive }) {
const [retainerTier, setRetainerTier] = React.useState('Standard');
const retainer = {
Essentials: {
price: '$3,500', per: '/ mo',
bullets: ['Quarterly audit refresh','Monthly Fix Cards','Async repair guidance'],
},
Standard: {
price: '$5,500', per: '/ mo',
bullets: ['Monthly audit','Fix Card delivery + implementation guidance','Biweekly working sessions','Hallucination monitoring','Narrative refresh'],
},
Strategic: {
price: '$9,500', per: '/ mo',
bullets: ['Weekly cadence','Founder-led implementation','Full-stack repair (content + technical + GitHub)','Priority access'],
},
};
const free = {
name: 'Free preview scan', tag: 'free · always',
price: '$0', sub: 'No credit card required',
blurb: 'See the obvious gaps before you invest in a deeper audit. Takes five minutes.',
bullets: ['Crawler-based preview scan','Surface-level crawlability, metadata, and technical signals','llms.txt and AI agent readiness signal','AI visibility scorecard at a glance'],
cta: 'Run your Free Scan',
};
const audit = {
name: 'AI Visibility Audit', tag: 'one-time · founder-delivered',
price: 'From $4,500', sub: 'Scoped to up to 25 priority URLs',
blurb: 'A complete AI visibility diagnostic and repair roadmap. Founder-delivered.',
bullets: ['Everything in the free scan','Deep hallucination scan vs. your fact sheet','Narrative alignment audit (6 dimensions)','Technical authority scoring across your domain','Buyer journey + source intelligence analysis','Content gap mapping','llms.txt + llms-full.txt generation','Fix Cards for content + engineering teams','60-min founder findings presentation'],
cta: 'Book an AI Visibility Audit',
};
return (
Go-to-market · Services
Start with a scan. Get the fix.
SurfaceGX delivers through founder-led engagements. Run a free scan to see where AI engines
are missing or misrepresenting your brand, then pick the depth of repair that fits.
Audit-to-retainer credit. Apply 50% of your full audit fee toward
your first month of retainer if you engage within 30 days of delivery.
Apply →
Also in motion
Self-serve · coming soon
SurfaceGX SaaS
Full access to the SurfaceGX diagnostic and repair stack as a self-serve platform. Currently in final development.
Agency partner program
White-label & co-sell
Built for technical SEO, content, and PR teams that need a credible path from monitoring report to shipped infrastructure fix.
);
}
/* ── Section: Get started ─────────────────────────────────────────────── */
function SectionStart({ setActive }) {
return (
Go-to-market · Get started
Where do you want to start?
Start with a free scan to see where AI engines are missing, misreading, or skipping your brand.
Use it to decide whether you need a fixed-scope audit or an ongoing managed program.
What the free scan checks
The preview scan gives you a fast read on the parts of the site that most often affect AI visibility.
✓Crawlability — whether the page can be reached cleanly by AI-facing crawlers.
✓Metadata and page-level signals — the signals that shape how the brand is interpreted by AI engines.
✓llms.txt and AI agent readiness: whether the files AI engines depend on are present and correctly structured.
✓AI visibility scorecard — the biggest gaps surfaced first, ready to share with your team.
If the free scan shows meaningful gaps, move into a fixed-scope AI Visibility Audit for deeper analysis or a Managed Program for ongoing remediation support.
Choose your path
Run a free Surface Audit
One URL. Full module-by-module score. Top 5 Fix Cards exportable as JSON or PR-ready
diffs. No account required for the first run.
Twenty minutes with Jeanine or John. They'll walk you through how the loop actually
runs on a portfolio like yours, and what'd ship in the first 30 days.
founder-led · ~20 min
);
}
/* ── Section: About ───────────────────────────────────────────────────── */
function SectionAbout({ setActive }) {
return (
Company · About SurfaceGX
Built to repair the infrastructure
AI engines read
SurfaceGX was founded by operators who saw the same pattern across brands: AI visibility
problems were being measured after the fact, but the machine-readable source layer was
still broken.
The founders
Two operators, one platform. Communications strategy meets data infrastructure.
Co-founder & CEO
Jeanine Morgan
Jeanine brings 20 years of expertise in communications, media relations, and content
strategy, having built and led PR and earned media programs for some of the most
reputation-sensitive brands in business.
She envisioned and co-founded SurfaceGX after experiencing firsthand how AI was
distorting brand narratives in ways no traditional communications measurement tool
could detect or explain.
John leads the product architecture behind the platform's audit, diagnostic, and
repair workflows.
John's background spans product development in AI, data infrastructure, enterprise
access layers, machine learning visualization, and hands-on product engineering. At
SurfaceGX, he focuses on making complex technical signals actionable for the teams
responsible for shipping the fix.
Most AI visibility platforms help teams monitor where a brand appears in generated
answers. SurfaceGX focuses on the layer before that answer exists: schema,
crawlability, canonical signals, source authority, page structure, and AI-readable
manifests.
POV · 02
The goal is practical repair.
Diagnose what AI systems cannot read, explain why it matters, and produce work that
content, SEO, and engineering teams can ship. No dashboards that only document
decline; every output is something a team can merge, publish, or hand off.
Start with a managed audit
See what AI systems are reading today.
Use a fixed-scope audit to see what AI systems are reading today, then decide whether
your team needs one-time remediation or an ongoing managed program.
Eight questions Marketing Leaders and Digital Marketing Teams ask before they invest in AI
visibility repair, answered without dashboard hand-waving. Each answer maps to the
SurfaceGX module that ships the fix.
FAQ content is only half the work. The other half is making sure engines can parse it
without ambiguity. This page emits FAQPage structured data. Here's what they ingest:
application/ld+json · FAQPageemitted on load
{`{
`}"@context"{`: `}"https://schema.org"{`,
`}"@type"{`: `}"FAQPage"{`,
`}"mainEntity"{`: [
{
`}"@type"{`: `}"Question"{`,
`}"name"{`: `}"How do I improve GEO performance for AI search?"{`,
`}"acceptedAnswer"{`: {
`}"@type"{`: `}"Answer"{`,
`}"text"{`: `}"Start by auditing whether your brand appears…"{`
}
},
`}// + 7 more questions, emitted from FAQ_DATA{`
]
}`}
setActive('concepts')}>
deeper
Definitions for GEO, AEO, the input layer, retrieval surfaces, and llms.txt,
the vocabulary these answers operate on, live in Key Concepts.
Key concepts →
Generative Engine Optimization (GEO) is the practice of making a brand consistently present, accurately described, and authoritatively cited by large language models when they generate responses about a category, topic, or problem.
Unlike traditional SEO, which focuses on ranking in a list of links, GEO focuses on being part of the synthesized narrative an AI produces. When someone asks ChatGPT "what are the best tools for X," GEO determines whether your brand appears in that paragraph, what it says about you, and whether it sounds authoritative or vague.
GEO is shaped by three factors:
Source authority Whether the sources that discuss your brand are high-quality, frequently cited, and varied.
Entity clarity Whether AI engines have a clear, consistent understanding of what your brand is, what it does, who leads it, and what makes it distinct.
Content structure Whether your pages are written and structured in ways that AI can extract, cite, and synthesize from.
Answer Engine Optimization (AEO) is the practice of structuring your content so that AI-powered answer engines (voice assistants, featured snippets, direct-answer interfaces) can extract your brand's information accurately and present it as a definitive response.
Where GEO focuses on brand presence in AI-generated narratives, AEO focuses on question-and-answer precision. AEO asks: when a buyer types "what does [Brand] do" or "how does [Brand] compare to X," can the AI produce a clear, factual, on-brand answer?
Strong AEO requires:
Clear heading hierarchy H1, H2, H3 structure that maps naturally to questions and answers.
Structured data Schema.org markup, especially FAQPage and HowTo, signaling content format to AI crawlers.
Authoritative bylines Author credentials and organizational affiliation that establish why your content is trustworthy on this topic.
Direct answer language Content written in plain, declarative sentences that answer specific questions without ambiguity.
llms.txt is an emerging way to make important website content easier for AI systems to understand. Placed at a domain root, it gives AI tools a concise map of the pages a brand wants represented clearly.
SurfaceGX helps teams prepare AI-readable manifests as part of a broader repair workflow.
llms.txt A concise index of priority pages.
llms-full.txt An expanded companion file for richer context when appropriate.
Both files are reviewed before deployment so the approved output reflects how the brand should be represented.
Tip
Manifest generation works best after the underlying pages have been cleaned up, clarified, and aligned with the brand's source of truth.
SurfaceGX is designed as a continuous improvement system, not a one-time report. Teams establish a baseline, identify the causes behind visibility gaps, prioritize the work, and measure whether changes improve AI understanding over time.
Detailed module-level workflows are available inside the authenticated SurfaceGX portal for approved customers and partners.
,
},
];
function SectionConcepts({ setActive }) {
return (
Resources · Key concepts
The vocabulary of AI brand visibility
Five concepts that anchor every SurfaceGX module. Understanding them helps you get more
from the platform and communicate findings clearly to your team.
setActive('modules')}>
apply these
Each concept maps to a SurfaceGX module. See Surface Audit, Engine Diagnosis,
Content Brief Empower, and Engineering Handoff in the Platform section.
Platform modules →
