Sources, standards, and limitations

Framework

Authorities we align to

Includes instructional-design references (cognitive load, multimedia learning, plain language, WCAG) used for owner deliverable presentation only — they do not change scoring formulas or outcome claims. Related owner terms: schema, NAP, NAICS, E-E-A-T, GEO.

Official standards and platform documentation

• AI Optimization Guide for Google Search (v2026-06-29) — Google Search Central. Primary reference for Google AI Overviews and AI Mode grounding.
• Apple Business Connect / Maps for businesses (v2026) — Apple. Claim and manage Apple Maps business listings; audit uses checklist + owner-supplied links only (no live scrape).
• Creating helpful, reliable, people-first content (v2026-06-29) — Google Search Central. Non-commodity, expert-led content for long-term visibility.
• Establish your business details on Google (v2026-06-29) — Google Search Central. Local business and product visibility via on-page details and GBP.
• Federal Plain Language Guidelines (v2011) — PLAIN / U.S. federal plain language program. Audience-first organization, short sentences, useful headings, active voice, lists over walls of prose. Owner copy standard for free UI, owner checklist, PDF, and limited packs. Companion: Digital.gov plain language guide series.
• Google Search Essentials (v2026-06-29) — Google Search Central. Baseline eligibility and quality expectations for Search indexing.
• Intro to structured data (v2026-06-29) — Google Search Central. Structured data supports rich results; not strictly required for generative AI.
• JavaScript SEO basics (v2026-06-29) — Google Search Central. Content should be available to crawlers without excessive JS dependency.
• North American Industry Classification System (NAICS) (v2022) — U.S. Census Bureau. Twenty major sector codes used for industry detection and workbook labeling.
• Page experience (v2026-06-29) — Google Search Central. Mobile usability and performance affect Search visibility.
• Plain language guide series (Digital.gov) (v2025) — U.S. General Services Administration (Digital.gov). Current federal digital plain-language guidance carrying forward plainlanguage.gov principles for web managers. Use for free in-page and hub HTML copy checks.
• Schema.org vocabulary (v2024) — Schema.org. Structured data types referenced in audit scoring and fix-guide paste blocks.
• Spam policies — scaled content abuse (v2026-06-29) — Google Search Central. Warns against thin doorway pages built only to capture query variants.
• Tips to improve your local ranking on Google (v2026) — Google Business Profile Help. Official local ranking factors: relevance, distance (proximity), and prominence.
• Web Content Accessibility Guidelines (WCAG) 2.2 (v2023) — W3C. Perceivable, operable, understandable, robust UI for free results and hub HTML. Focus-visible, labels on checkboxes, contrast, and keyboard access for prove-it and step accordions — part of delightful without excluding.

Research and methodology references

• 10 Usability Heuristics for User Interface Design (v2024) — Nielsen Norman Group. Visibility of system status, match real world, user control, consistency, error prevention/recovery, recognition over recall, aesthetic minimalism, help. Applied to free/in-page and hub launcher status lines and plain-language recovery.
• A Theory for Record Linkage (v1969) — Journal of the American Statistical Association. Fellegi & Sunter foundational probabilistic record linkage — basis for multi-source NAP agreement (on-site schema vs visible; owner-attested GBP vs site). Not a ranking model.
• AI Overview and citation rate analysis (v2026) — Ahrefs. Industry benchmark framing for E-E-A-T bylines and sourced claims; used as directional context only.
• Cognitive Load Theory (v2011) — Springer (Sweller, Ayres & Kalyuga). Foundational cognitive load theory: manage intrinsic load, minimize extraneous load, free germane load for schema building. Applied to owner deliverables as one-job-per-screen, de-duplicated disclaimers, and deferred technical detail.
• Competing Against Luck — Jobs to Be Done (v2016) — Clayton M. Christensen / Christensen Institute. Customers 'hire' products for progress. Free pack job: see and fix top easy gaps. Mid job: finish the next high-priority checklist pack (≥5 when available). Full job: complete scorecard + remaining work + implement hub.
• F-Shaped Pattern for Reading Web Content (eyetracking) (v2017) — Nielsen Norman Group. Users scan; front-load conclusions and action labels on the first horizontal pass and left edge. Owner pages put score + next action above the fold; avoid burying the first fix under long disclaimers.
• First Principles of Instruction (v2002) — Educational Technology Research & Development (M. David Merrill). Task-centered instruction: activation, demonstration, application, integration. Each easy-fix card is a real owner task with demonstration (payload), application (publish + checkbox), integration (re-audit / full pack).
• Fogg Behavior Model (B=MAP) — motivation, ability, prompt (v2019) — BJ Fogg / Behavior Design Lab (Stanford). Behavior happens when motivation, ability, and a prompt converge. Free path raises ability (paste-ready steps) and motivation (clear score + progress), then places one clear mid/full CTA prompt after value — not before results.
• GEO: Generative Engine Optimization (KDD 2024) (v2024) — ACM KDD / arXiv:2311.09735. Aggarwal et al.: generative engines synthesize multi-source answers; GEO methods can lift visibility up to ~40% on GEO-bench. Entity-consistent local facts (NAP/hours) support citation readiness — not a traffic guarantee.
• GeoSummit.ai Framework (v0.4.7) — Audit methodology (this tool). v0.4.7: twenty NAICS sector profiles, eighty SMB subcode packs, finding catalog, CMS playbook, owner PDF + developer snippet exports.
• Influence — The Psychology of Persuasion (principles of influence) (v2021) — Robert B. Cialdini / Influence at Work. Ethical application of reciprocity (free value first), commitment/consistency (prove-it after free fixes), authority (methodology link), and honest scarcity (real remaining checklist work only — never fake countdown timers). No dark patterns; Digital Visibility Score readiness only.
• Multimedia Learning (Cognitive Theory of Multimedia Learning) (v2009) — Cambridge University Press (Richard E. Mayer). CTML principles used for pack presentation: coherence (cut noise), signaling (kickers/step numbers), segmenting (Top-3 then rest), spatial contiguity (instruction beside payload), pre-training (what the score is before deep steps).
• Nudge — Improving Decisions about Health, Wealth, and Happiness (v2008) — Thaler & Sunstein. Choice architecture: default the helpful path without coercion. Soft CTA after free value, clear price, opt-in email, human publish retained — libertarian paternalism, not hard modals or dark patterns.
• Principles of Instructional Design — Nine Events of Instruction (v1992) — Gagné, Briggs & Wager / NIU CITL summary. Classic nine events: gain attention → inform objective → present content → guided practice → feedback → assess → enhance retention. Owner journey: hero status → time-to-first-green goal → numbered fixes → mark published → re-audit.
• Prospect Theory — An Analysis of Decision under Risk (v1979) — Econometrica (Kahneman & Tversky). Loss aversion: people weigh open gaps more than abstract gains. Owner CTA names real unfinished checklist items still public on their site — without claiming traffic/lead loss. No guaranteed outcomes.
• Schema markup and AI search traffic studies (v2025) — Search Engine Land. FAQ and video schema associated with AI referral lift in some vertical analyses.
• Structured data and generative search visibility (SSRN working papers) (v2024) — SSRN. Academic/industry literature on structured author and entity signals for AI attribution.
• The Adult Learner — Andragogy in Practice (v2020) — Routledge (Knowles, Holton, Swanson). Adult learners are problem-centered, experience-driven, and self-directed. SMB owners get immediate real tasks (fix my site now), not abstract SEO lectures; optional depth for developers/agents via progressive disclosure.
• e-Learning and the Science of Instruction (v2024) — Wiley (Ruth C. Clark & Richard E. Mayer). Evidence-based e-learning guidelines (worked examples, practice + feedback, segmenting). Maps to free prove-it checkoffs, re-audit steps, and paste-ready worked examples without guaranteeing business outcomes.

Industry benchmarks (directional context only)

• Consumer local search behavior (industry research) (v2025) — BrightLocal. Industry survey: map apps account for a material share of local discovery. Directional consumer research — not a peer-reviewed RCT.
• Don't Make Me Think, Revisited — A Common Sense Approach to Web Usability (v2014) — New Riders (Steve Krug). Self-evident navigation and 'billboard' clarity under time pressure. Owner WOW test: can a busy SMB decide the next 15-minute action without studying?
• Third-party AI citation-readiness model (v2.3) — Audit methodology (this tool). Offline on-page signal scoring for industry shopper questions; live AI validation is manual (paste into ChatGPT/Claude).

Validation and map attribution tools

• Geoapify Static Maps API (v2026) — Geoapify. Primary basemap for competitor proximity maps when GEOAPIFY_API_KEY is set. Free plan allows commercial/print use with required attribution (Powered by Geoapify).
• Google Rich Results Test (v2026) — Google Search Central. Post-publish rich-result eligibility check linked from fix guides.
• OpenMapTiles (v2026) — OpenMapTiles. Vector tile schema used by Geoapify map styles (e.g. osm-bright). Attribution required with Geoapify-rendered static maps.
• OpenStreetMap (v2026) — OpenStreetMap contributors. Map data and attribution for competitor proximity maps (Phase 0 OSM basemap and Geoapify path).
• Schema.org Validator (v2026) — Schema.org. Post-publish JSON-LD validation linked from fix guides.

Deprecated or limited policies

• Changes to HowTo and FAQ rich results (v2023-08-08) — Google Search Central Blog. FAQ rich results limited to government/health; not a retail visibility tactic.

Certification

ISO certification pursuit — not yet certified

Building toward ISO 9001:2015 and ISO 27001:2022 with documented QMS/ISMS packs. Third-party certification is pending — no certificate issued yet. Readiness is not the same as certified.

• ISO 9001:2015 — quality management (alignment target)
• ISO 27001:2022 — information security (alignment target)

Company — Standards & Innovation →

GeoSummit methodology

Internal framework and white papers

Three cross-referenced documents behind the pipeline, backlog governance, and self-audit — with stated scope and non-claims. Methodology v0.4.7 · content v1.0.118.

1. GEO/AEO Audit white paper — Pipeline architecture and limitations. Download → · terms: GEO, export gate, sector-weighted
2. Product Backlog white paper — Backlog command center analysis. Download →
3. Website self-audit white paper — We run the same audit on our site (built for 95+; 93 then multipage pack 98). Download → · Internal proof on Proof

We separate what automation does for customers from how AI accelerates product development — see Company, Methodology, and the Glossary (GEO, AEO, RAG).

Full methodology → · Construct validity → · Is this really AI? → · Glossary →