The 5-signal framework that decides whether ChatGPT, Perplexity, and Google AI Overviews recommend your business — or hand the customer to a competitor. Built for founders, marketers, and AEO/GEO practitioners who refuse to be invisible where buyers actually ask.
Get a free inside look at “Foundations: How AI Engines Decide Who to Recommend (AEO, GEO & the 4-Layer Search Reality)” — the video overview.
By the numbers
A real customer, budget in hand, asked ChatGPT who to hire — and it answered confidently, in full sentences, with three businesses. Yours wasn't one of them. Unlike a results page, there is no page two. Or Perplexity cited a competitor's three-year-old blog post as the authority in your category. Or Google's AI Overview answered the question above your #1 ranking and the click never happened. Worst of all, none of it shows up in your analytics. Now multiply that by every buying question in your market, every day — you're not just losing clicks, you're losing the recommendation itself, invisibly, at velocity.
Five signals answer engines weigh when they decide which businesses to name — measured, strengthened, and compounding. S¹ Business Identity: become an unambiguous entity AI can confidently name. S² Review Strength: build the depth, velocity, and recency models read as proof. S³ Website Content: publish answer-shaped pages AI quotes verbatim. S⁴ Structured Data & Schema: declare who you are in the language machines parse. S⁵ External Citations: earn mentions on the sources AI already trusts. Each signal maps to a concrete set of fixes you can audit, implement, and verify on your own business — not a sandbox.
Module 1 builds the foundations — AEO, GEO, and how AI decides — and has you measure a baseline AI citation share. Each of the next five owns one signal, ends with a hands-on workshop, and produces a measurable visibility improvement. M2 fixes identity and entity conflicts. M3 builds a compliant review engine. M4 restructures pages into answer-shaped, citable content. M5 deploys and validates copy-paste JSON-LD. M6 earns external citations and closes with your full S⁵ audit and a 90-day roadmap.
Founders and local businesses who don't have an SEO team — they are the SEO team — get a finite, ordered checklist that makes AI engines name them when buyers ask, with no agency retainer. Marketers and SEO leads already ranking but watching AI answers eat the clicks above their listings get the framework to defend that demand and the vocabulary to brief their team, boss, and board on what comes after the ten blue links. AEO/GEO practitioners get a repeatable system to operationalize.
Three months from now the answers are already being written without you. Six months from now AI assistants have settled on a shortlist for your category — a handful of entities they recognize, trust, and repeat — and re-asking the question keeps returning the same competitors. A year from now the share of buying journeys that start in an AI chat has doubled again, and your beautifully ranked website is a storefront on a street nobody walks down. The businesses that installed the signals early are getting named first — on the same platforms you have access to right now.
AI is now the third most-used local discovery channel — ahead of Yelp and TripAdvisor — with ChatGPT the frontrunner (used by 31% of consumers for business recommendations in the past 12 months).
50% of consumers have made a purchase after using AI during research and 69% expect AI to play a bigger role in how they shop, per Semrush's 2026 buyer-journey study.
6 modules, each ending in a deliverable.
S⁵ is the five AI visibility signals answer engines weigh when deciding which businesses to name: S¹ Business Identity, S² Review Strength, S³ Website Content, S⁴ Structured Data & Schema, and S⁵ External Citations. S⁰ (Foundations) is the orientation layer. Every fix you'll ever make to your AI visibility lives under exactly one of these pillars.
AEO (Answer Engine Optimization) targets being the single answer for voice assistants and featured snippets; GEO (Generative Engine Optimization) targets being named in Perplexity, ChatGPT, and Claude answers; classic SEO targets Google rankings and clicks. They overlap but reward different work — which is why a single framework that feeds all four search layers is more efficient than chasing each separately.
The default set is Perplexity, ChatGPT in search mode, Claude, and Google AI Overviews. Running your priority buyer queries through all four gives you a representative citation-share baseline, since these are the surfaces where buying questions are actually asked.
Citation share is the percentage of your priority buyer queries where an AI engine names or cites your business: (checks where you appear ÷ total checks) × 100. It's the new KPI because rankings have become a lagging indicator — you can rank #1 and still be absent from the answer that wins the customer.
Not directly, on its own. The largest controlled study (Ahrefs, 1,885 pages) found no statistically significant citation uplift from adding JSON-LD. Schema is best understood as entity-clarity hygiene and an amplifier for pages that already have authority and clear content — valuable when accurate and rich, but not a substitute for citations and clear writing.
Three things, in order: earned branded web mentions across third-party sources (the strongest single predictor, ~0.664 correlation), content that is fresh and answer-first (associated with ~2.8× higher citation rates), and consistent brand positioning across the web. Roughly 85% of AI brand mentions originate off your own site.
Because every other signal attaches to your entity. If your name, address, and phone conflict across listings, or AI confuses you with a same-named business, the engine hedges — and hedging engines don't cite you. Clean identity is the foundation reviews, content, schema, and citations all build on.
Very — models read reviews as proof you're the safe recommendation, weighting velocity, recency, and breadth over raw star count. With 45% of consumers now asking AI for local recommendations and AI trusted as much as reviews by 42% of them, your review engine directly affects whether you make the AI shortlist.
A real buyer question as the heading, a direct 40–80 word answer immediately below it that names your business, and depth after. A page that explicitly states 'the best ___ for ___ is [your business] because [reasons]' in plain text gets quoted; a page that never makes a direct claim gets skimmed past, no matter how good its schema is.
Start with a lightweight, repeatable monthly playbook — a target list across the six source types (directories, trade press, podcasts, partner sites, news, associations), pitch templates, and one owner — aiming for 5–10 quality citations a quarter. The cheapest next citation is usually a source type where you currently have zero.
Some signals move fast: identity can be cleaned in about a week and schema is same-day copy-paste. Reviews and content compound over weeks. Citations are the slowest — and that's exactly why they compound for whoever starts first. The 90-day capstone roadmap sequences the work so you lift your lowest signal first and re-score quarterly.
Both. The signals are universal — identity, reviews, content, schema, citations — though the emphasis shifts: local and service businesses lean hard on Google Business Profile and review breadth, while B2B and online businesses lean on editorial citations and answer-shaped content. The framework adapts; the sequence rule (lift your lowest signal first) is the same.
The surfaces change; the underlying decision doesn't. Every engine still asks: do I recognize this entity, is it the safe choice, can I quote it, can I parse it, do trusted sources vouch for it? S⁵ maps to those durable questions, not to any one platform's current UI — which is why the framework survives the next interface change.
A real, dated AI citation-share baseline for your own business, an understanding of how AI decides who to cite, a competitive citation map of who's winning your top query, and a 0–5 score across all five signals so you know exactly which pillar to lift first — all in under 90 minutes.
Treat it as a system, not a project: track citation share weekly across the four engines, run a quarterly content refresh, maintain a monthly citation-retention check (target 90%+ retained year over year), and re-score all five signals each quarter to find your new lowest. The architects defend and compound; the rest let it decay.
S⁵ is a single-edition, hands-on program: six modules, thirty lessons, one framework, applied to your real business. You leave with a measured baseline, a clean entity, citable content, validated schema, a citation playbook, and a 90-day roadmap. The answers are being written right now — make sure yours is in them.
