Playbooks

Step-by-step, self-contained playbooks.

• AI Citation Tracking

  Use a repeatable manual and automated workflow to measure how often, how prominently, and on which AI engines your content is cited. Define the prompt set, sample answers, verify URLs, normalize the data, and report results using the definitions in GEO Metrics.

• AI Crawler Access Audit

  An AI crawler access audit compares four forms of evidence: intended policy, live declarations, actual responses, and verified arrivals. Differences between adjacent states reveal access failures, including network-level blocks that robots.txt cannot show.

• Brand Mention Tracking

  Track brand mentions in AI answers by running an alias-aware detector against a frozen prompt set, verifying each match, and calculating Mention Frequency, Share of Voice, Answer Inclusion Rate, and Brand Sentiment as defined in GEO Metrics.

• Citability Audit

  A citability audit tests whether individual passages can stand alone when an AI system retrieves them for an answer. Start with a manual chunk-extraction test, assess seven signals, rate each finding by severity, and record the appropriate rewrite for each failure.

• Deploying llms.txt

  Deploy llms.txt in four stages: curate the file, generate it from site data, and serve clean markdown for its links. No major vendor confirms consumption and no official validator exists, so limit the initial work to about a day and remove an unmaintained file.

• Full GEO Audit

  A full GEO audit reviews six dependent areas in order: access, rendering, structure, content, off-site authority, and observed outcomes. This playbook explains how to sequence the checks, decide when later checks are meaningful, assign severity, and produce an actionable report.

• GEO Maturity Model

  Assess GEO capability across five dimensions and five levels: Unmanaged, Instrumented, Systematic, Competitive, and Reference. Your overall level is the lowest dimension score, not the average. Each level has an exit test, priority actions, relevant KPIs, and a common pitfall.

• Schema Audit

  A schema audit tests two consumers of inherited markup with opposite failure modes: parsers silently discard invalid blocks, while live-fetch models still read them as page text. It checks coverage, validity, integrity, and truth; only truth failures carry a documented penalty.

• Schema Implementation

  Deploy Schema.org in three tiers: entity identity first, page types second, and conditional markup for assets and answer formats last. Render it in the initial HTML, run four checks, and keep it consistent with the visible page. Expect easier parsing, not more citations.

• Writing for AI Citation

  Use seven signal-specific recipes to make passages easier for AI systems to cite. Each recipe includes a before-and-after example, and the playbook adds an MDX template, a pre-publish checklist, and common mistakes to avoid.