Spearleaf · Position Zero Playbook v10 · 2026-06-16
Start here Changelog
Strategy

Winning AI / LLM Search Visibility

Google ranks pages, AI ranks entities; earned signals teach the models; consensus beats raw authority.

On this page

This is the Playbook's page on getting your brand named, cited, and recommended inside AI answers. It synthesizes where a dozen experts at the 2026 conference independently landed: Google ranks pages, but AI ranks entities, and the earned signals that build entity consensus are what teach the models to cite you. Read it for the strategy and the through-line, then jump to the linked system pages to actually execute.

AI and LLM search visibility was the defining anxiety and opportunity of SEO Spring Training 2026. Across all three days, speakers kept returning to one shift: the job is no longer (only) ranking a page in ten blue links, it is being named, cited, or recommended inside an AI answer from ChatGPT, Claude, Gemini, Grok, Perplexity, and Google's AI Overviews / AI Mode. The room agreed the old keyword-and-rankings playbook is stale, that LLMs reason over entities and consensus rather than pages, and that off-page authority (earned media, referring domains, reviews, social) now feeds the models as training and citation fuel. Where speakers diverged was on the mechanism (file engineering vs compute eligibility vs earned consensus vs referring-domain diversity) and on how clean or gray-hat the path to visibility should be. This brief synthesizes the nine AI-visibility sessions. For the off-page link and local mechanics underneath much of it, see the Spearleaf getting-cited-by-AI page and the off-page anchor system.

The through-line

Five ideas that multiple speakers, independently, converged on:

  1. Google ranks pages; AI ranks entities (and you must define your entity consistently everywhere). Dawood Bukhari stated it most directly (Google ranks URLs, keywords, backlinks; AI understands brands, maps relationships, evaluates credibility, cites sources) and built his NARC method around answering "who you are, who you serve, what you solve" consistently on every surface. Daryl Osborne (Surge Protocol) made the same page-to-entity reframe ("for ten years we asked how do I rank; the question now is am I in the answer?") and built a measurement tool around AI citation. Brian Winum framed the whole thing as E-E-A-T plus "Brand. Authority. Data." mapping to entity authority. Kyle Roof reframed the same goal as "become your own source" inside ChatGPT.

  2. Off-page earned signals, not on-page polish, are what teach the models. Dawood Bukhari cited that 85% of AI brand mentions come from earned media (not brand-owned content) and put digital PR at the literal center of his "AI consensus ecosystem." Brian Winum manufactured the same signal supply (listicles, roundups, advertorials, research reports, PR) as authority fuel. Ted Kubaitis reframed referring domains as the connective signal across channels that Google (and by extension AI) reads. Michael Merlino argued social activity is now how Google reads trust and prominence. Each is a different on-ramp to the same conclusion: presence across many third-party surfaces is the engine.

  3. Consensus and repetition beat raw authority. Dawood Bukhari's NARC (Narrative consistency, Authority density, Repetition, Co-occurrence) argues a lower-DR brand doing consensus work out-ranks a higher-DR brand that is not, because "AI doesn't want to be wrong" and repeats what 500 sources agree on. Brian Winum's "authority stacking" (repeated appearances in roundups, listicles, PBNs) is the same repetition-as-consensus logic. Surge's "Collaborative" and "Amplifying" dataset roles describe the same phenomenon from the engine's side: multiple independent sources corroborating you is what earns the citation.

  4. You can win AI visibility from a weak traditional footprint, which is the opening. Surge (Daryl Osborne) showed a free Weebly subdomain with no backlinks cited alongside premium grill brands, and a manufacturer surfacing for a local buying-intent query. Kyle Roof's term-selection method wins terms "Google already trusts you for" with minimal SEO and then satisfies LLM query fan-out. Dawood Bukhari's Cabo villa case (brand-new business, no site) and Eldar's whole pitch frame this as the reason to move now: brands dominant on Google but weak in LLMs can be beaten on the AI game.

  5. Local angles and local entities are a disproportionate AI-visibility lever. Dawood Bukhari found local angles get picked up most (one national dataset spun into ten or fifty localized journalist stories, earning local TV mentions). Ted Kubaitis showed a referring-domain footprint betrays its city and can be pushed to the neighborhood level. Eleftherios Livadaras (Elias) built his entire local hierarchy (relevance, proximity, prominence) toward greening a geo grid. Michael Merlino's UGC video play is deliberately geo-targeted (PAA plus geo).

Tactics & playbook

Concrete, do-this items pulled from the talks:

Tensions & disagreements

Sources (conference sessions)

These are conference session references, not pages on this site. Speakers and talk titles, as named in the source notes:

Connect it to your system

Turn the synthesis above into running systems with these Position Zero pages:

New here? Start with the overview.