The reading library.

Numerator / denominator: Valid observations with an observed subject-domain supporting link / eligible valid observations. Review rule: Count each observation once, even if it cites several subject URLs.

Read the source guide: 12 AI Visibility Metrics That Matter More Than a Single Score →

Calculation: Valid recorded observations / recorded attempts, with retries identified. Interpretation: Collection completion; errors and unavailable outputs are not brand absences.

Read the source guide: 12 AI Visibility Metrics That Matter More Than a Single Score →

Calculation: Correct checked claims / (correct + incorrect checked claims). Required companion information: Show unverifiable claim counts and category breakdown separately.

Read the source guide: 12 AI Visibility Metrics That Matter More Than a Single Score →

Keep them separate for decisions. Any combined score introduces weighting choices and can hide missing data, inaccurate descriptions or weak inquiry quality behind a rising average.

Read the source guide: 12 AI Visibility Metrics That Matter More Than a Single Score →

These measures come from analytics and qualification records, not from a prompt panel. OpenAI documents ChatGPT referral tagging ; this helps identify some inbound traffic but does not reveal every prior AI interaction or the visitor’s original prompt. Use a documented source rule and a consistent period. In GA4 Traffic acquisition, session-scoped dimensions support a session-level view. Keep self-reported discovery and separately attributed CRM activity alongside it, without adding overlapping counts.

Read the source guide: 12 AI Visibility Metrics That Matter More Than a Single Score →

Collection quality comes first. A percentage from three successful answers should not look as complete as one from a fully observed panel. Keep failed attempts visible and prevent retries from becoming extra independent successes. For prompt coverage, choose the eligible panel group before collecting. Ten identity questions and sixty non-branded questions answer different questions and must not share a coverage denominator.

Read the source guide: 12 AI Visibility Metrics That Matter More Than a Single Score →

A confirmed mention identifies the intended company, not an unrelated business with the same name. A source citation is an observed supporting link to the subject domain. A recommendation explicitly proposes the company for the buying need. Review each independently. For a valid Google search with no AI Overview, preserve that outcome in overall search-observation rates and separately report whether an Overview appeared. A rate conditional on triggered Overviews needs a different, clearly named denominator.

Read the source guide: 12 AI Visibility Metrics That Matter More Than a Single Score →

Fact accuracy applies to checked claims, not entire answers. Archive the applicable pricing, feature or policy reference with its date and version. Classify unsupported claims as unverifiable rather than silently treating them as correct or incorrect. Repeat agreement describes stability, not truth. Compare the same questions, contexts and classification rule across independent sessions. A brand can be consistently absent or consistently misdescribed. Do not interpret high agreement as high visibility.

Read the source guide: 12 AI Visibility Metrics That Matter More Than a Single Score →

Test buyer questions that include the actual problem and desired deliverable. Save answer context and inspect whether citations support the expertise and scope described. Distinguish a mention of an expert from a recommendation of the current firm. In discovery records, note whether buyers understood deliverables, ownership and exclusions before the meeting. Track qualified briefs and recurring misunderstandings alongside observed citations. Use the citation tracking guide to keep the measurement definitions explicit, and the free audit to choose the first source gap to address.

Read the source guide: A Consulting FAQ System That Explains Expertise, Deliverables, and Fit →

Example: A surface has an estimated visibility result only. Acceptance condition: The result is labeled correctly and missing observations remain explicit.

Read the source guide: A GEO Audit Playbook: From Findings to Fixes →

Prioritize a factual correction before an optional new article. Among remaining tasks, consider buyer importance, strength of the evidence, effort and whether the proposed asset would help even without an AI citation. Give each selected task an acceptance condition. 'Publish current onboarding dependencies with product-owner approval' is checkable. 'Increase authority' is too vague. Record a publication date and the questions whose answers may be relevant to review. Keep a hypothesis separate from an observed cause. A useful source gap can justify improving information, but it does not prove that a particular edit will make an assistant recommend the company.

Read the source guide: A Weekly Citation Tracking Playbook →

Start with commercially important questions, product inaccuracies and large unexplained changes. Open the saved responses. Check whether a new source supports a recommendation, an educational claim or a criticism. The same domain can play different roles. Review a sample of apparently unchanged answers as well. A stable total can conceal different cited pages, a new competitor or an inaccurate product description. Retain the full context rather than extracting only the sentence that mentions your company. Bing AI Performance offers citation observations across supported experiences. Use it as complementary evidence, keeping its coverage separate from your own prompt panel. Bing's report introduction.

Read the source guide: A Weekly Citation Tracking Playbook →

A source gap is a documented information need worth investigating. It might be a missing implementation example, an outdated price explanation or a relevant independent comparison where the product is absent. Record the evidence before proposing an action. Each journal entry should include the question ID, response reference, source URL, date, affected buyer need, proposed owner and unresolved uncertainty. Do not label every competitor citation an opportunity: the source may cover a different product category or market. Separate changes you control from external opportunities. Correcting your documentation can be assigned immediately within your publishing process. An external mention depends on relevance and another publisher's independent decision.

Read the source guide: A Weekly Citation Tracking Playbook →

Investigation: Find the source and verify the current limit. Possible work item: Correct the owned page and review related material.

Read the source guide: A Weekly Citation Tracking Playbook →

No. The review checks available observations, investigates important changes and creates work items. The collection schedule should be defined separately and remain comparable. Not every question needs to be recollected during every review.

Read the source guide: A Weekly Citation Tracking Playbook →

No. First check audience, category, accuracy and editorial relevance. Some sources are unsuitable or outside your influence. Record the rationale before preparing an authorized contribution.

Read the source guide: A Weekly Citation Tracking Playbook →

Include collection quality, important findings, completed corrections, unresolved gaps, owners and the next review date. Link each finding to its evidence and retain negative or inconclusive outcomes.

Read the source guide: A Weekly Citation Tracking Playbook →

Confirm completed changes on their public URLs and verify the claims, links and next step. On the next comparable collection, review the relevant questions without removing poor results. Note whether the issue remains, improves or cannot yet be evaluated. An operational summary should list completed corrections, unresolved evidence gaps and the next collection date. Retain examples that produced no measurable change; they help prevent repetitive work based on an attractive assumption. Apply the journal to the healthcare citation dossier, where claim context requires careful review, or the yacht-broker visibility dossier, where buyer fit matters. Request a free audit to identify a practical starting scope.

Read the source guide: A Weekly Citation Tracking Playbook →

Collect the current observation export, the previous comparable period and the page-change log. Keep the question-panel version and collection context visible. Before interpreting a chart, check whether the same questions and surfaces were actually measured. Identify incomplete attempts and settings changes. If a tool switched from a public interface to an API, separate that series. If the sample shrank, show which questions are missing. A recording problem should create a measurement task before it becomes a content recommendation. Use the reference measurement method for classification and denominators. This weekly process adds an operational journal; it should not create a competing definition of citation rate.

Read the source guide: A Weekly Citation Tracking Playbook →

The implementation task is complete when its stated acceptance condition is verified. A later visibility outcome is a separate observation. Keep both statuses so a published correction is not mistaken for a proven recommendation gain.

Read the source guide: A Weekly Citation Tracking Playbook →

Measure indexed pages, prompt visibility, citations, referral clicks, branded search, form submissions, demo requests, CRM notes, and pipeline influenced by answer-ready content.

Read the source guide: AEO Checklist for Answer-Ready Pages →

Record it as evidence from the specific API configuration used. Do not label it as an observation of the public ChatGPT Search interface unless that interface was actually tested. Keep different collection environments in separate panels.

Read the source guide: AI Citation Tracking: A Reproducible Measurement Method →

No. A citation can support an educational statement while the answer recommends another company. A recommendation must explicitly present your company as a solution to the buying need, with its identity verified.

Read the source guide: AI Citation Tracking: A Reproducible Measurement Method →

Use this page as the common definition of evidence and outcome rates. Each guide below owns a different decision; the measurement collection groups all fourteen guides into four reading paths.

Read the source guide: AI Citation Tracking: A Reproducible Measurement Method →

An observation is one attempted question on one specified surface in one session. Record the attempt even when it fails. A valid observation is a completed result under the defined collection conditions; technical failures remain excluded from outcome-rate denominators and are reported separately. Create commercial questions about choosing a solution and informational questions about understanding a topic. Keep company-named identity checks in a third cohort. Recommendations prompted by your own brand name must not inflate the commercial discovery rate. Preserve the wording, question ID and panel version. Adding a question changes the sample, so report additions separately from the original trend.

Read the source guide: AI Citation Tracking: A Reproducible Measurement Method →

Record the public interface or API configuration, search mode, language, known location settings, timestamp and session context. Use consistent conditions for repeated comparisons and record unavailable details as unknown. A prompt's requested country does not prove the actual search location. An API answer is evidence about that configured API call, not automatically the corresponding consumer search product. Similarly, Google AI Mode and AI Overviews are separate surfaces. Keep their results separate even when the questions overlap. Schedule repetitions across several days. Define retry rules before collecting. Retain the failed attempt and mark its replacement so that a retry does not become an extra success. Multiple completions of the same question are repeated observations, not independent buyers.

Read the source guide: AI Citation Tracking: A Reproducible Measurement Method →

Maintain a page-change log and keep the original question cohort available. Compare question-level outcomes across sessions before interpreting an aggregate increase. A change in answer behavior can have several causes beyond your intervention. Use the benchmark protocol for a comparison study and the 90-day review guide for decisions. Start with a free audit to identify which questions and site checks are relevant.

Read the source guide: AI Citation Tracking: A Reproducible Measurement Method →

Domain citation rate equals valid observations with at least one qualifying domain citation divided by valid observations in the stated cohort. Count an answer once even when it links to several pages on that domain. Use a separate distinct-URL table to explore which pages receive citations. Commercial recommendation rate equals valid commercial observations recommending the company divided by all valid commercial observations in that cohort. A negative mention, incidental reference or informational citation is not a commercial recommendation. For AI Overviews, a valid search without an Overview remains in the overall denominator with no Overview citation or recommendation. Also report trigger rate and conditional outcomes among triggered Overviews. An execution error is different and remains outside these rates. Report n/N, not only percentages. A top-three measure applies only to explicitly ordered recommendation lists, with its eligible-list count shown. Do not invent a rank for unranked prose.

Read the source guide: AI Citation Tracking: A Reproducible Measurement Method →

Google's Generative AI performance report shows impressions in covered Search features, with page, country, date and device views. It does not directly report whether a brand was recommended. Search Console documentation. Bing AI Performance reports citations across its supported experiences. Keep that platform-provided evidence separate from your custom prompt panel. Bing documentation. Link answer observations to attributable site visits when evidence permits, then follow audits, qualified inquiries and customers. Retain self-reported discovery information. Avoid assigning every direct visit or subsequent sale to AI.

Read the source guide: AI Citation Tracking: A Reproducible Measurement Method →

Version 2 contains 60 non-branded questions about GEO and AI visibility solutions: 40 commercial and 20 informational questions, plus ten separate Arrow identity controls. Use the commercial questions to assess solution discovery in the stated contexts; exclude the identity controls from that recommendation rate. The questions are a research starting point, not verified search demand. Download the version 2 question panel, the observation contract and the action journal contract. The kit instructions explain evidence requirements, missing outcomes, repeat sessions and comparable before-and-after periods. The provided observation journal and action journal are empty. Collect and review real answers before producing a baseline. Store completed records privately; do not publish customer or session data as part of a public template. The earlier version 1 panel remains available for separate downstream research, including yacht-brokerage and DPE-provider questions. Those subjects and their recommendations must not be merged into the version 2 GEO-solution discovery denominator.

Read the source guide: AI Citation Tracking: A Reproducible Measurement Method →

Include valid searches without an Overview in the overall denominator as having no Overview citation or recommendation. Also report trigger rate and conditional outcomes among triggered Overviews. Exclude and separately report technical execution failures.

Read the source guide: AI Citation Tracking: A Reproducible Measurement Method →

Counts when: The company is explicitly proposed for the buying need. Does not establish: First place or exclusive preference.

Read the source guide: AI Citation Tracking: A Reproducible Measurement Method →

Counts when: A visible supporting source links to the company's domain. Does not establish: That the company is recommended.

Read the source guide: AI Citation Tracking: A Reproducible Measurement Method →

Counts when: A factual claim matches a verified current product source. Does not establish: That all other claims are correct.

Read the source guide: AI Citation Tracking: A Reproducible Measurement Method →

No. Readiness checks describe the website under a defined rubric. Citation measurement requires an observed answer with a qualifying source link. Reports should keep those results separate and identify estimates explicitly.

Read the source guide: AI Citation Tracking: A Reproducible Measurement Method →

Check the company's identity using its domain and context. A shared name alone is insufficient. Preserve ambiguous cases for review rather than assigning them silently. Use these definitions consistently across every surface.

Read the source guide: AI Citation Tracking: A Reproducible Measurement Method →

Keep the complete answer, visible sources, observation ID and classification together. Record an actual shared-result link or capture when available. Avoid replacing the evidence with a generated summary that cannot be checked. Mark whether a source was visibly cited. Do not claim a page was retrieved merely because it seems relevant: retrieval is an additional event that may not be exposed. Record it only when the collection system provides that evidence. Review ambiguous brand matches, recommendations and factual claims manually. Recheck a sample of ordinary classifications and document corrections. Store the original classification alongside its revision so that a chart can be regenerated.

Read the source guide: AI Citation Tracking: A Reproducible Measurement Method →

The following is a fictional arithmetic example, not an Arrow result or an industry benchmark. Assume one commercial question cohort, one public interface and a frozen panel. There are 30 planned attempts: 27 completed answers and three technical failures. The completed answers contain nine brand mentions, six domain citations and four explicit recommendations. These outcomes can overlap. The mention rate is 9/27, the citation rate is 6/27 and the recommendation rate is 4/27. Three failures are reported as 3/30 attempts; they are not treated as valid non-mentions. Counting several links in one answer does not create additional cited answers. If there were no valid answers, the outcome rates would be not measured. This small panel describes its collected answers. It does not estimate all users, prove improvement or identify a causal effect from a content change. Add the question coverage and session breakdown before comparing another period.

Read the source guide: AI Citation Tracking: A Reproducible Measurement Method →

Yes, as a baseline for future work. It is not a before snapshot for the change already made, and it cannot establish that change’s effect.

Read the source guide: AI Citation Tracking: What to Measure Before You Change Content →

Start with one observable problem. “The assistant describes our annual plan as monthly” suggests a pricing-clarity correction. “We want more visibility” does not identify which page, fact or buyer decision should change. Write the expected reader benefit even if no assistant changes its answer. Then separate correction from experimentation. Fix a demonstrably incorrect public price or broken link promptly. Do not keep harmful or misleading content live to preserve an experiment. If there is no time to capture an AI baseline, record that limitation and treat the work as a verified website correction.

Read the source guide: AI Citation Tracking: What to Measure Before You Change Content →

One action entry should connect a buyer question to a page, the observed problem and the actual patch. “Optimized for GEO” is too vague to review. “Added an annual billing label beside the displayed price and clarified the cancellation paragraph” identifies an intervention. Use a sequence of planned, implemented, retested and closed. Implementation requires a dated patch or deployment reference. Retesting requires real later observations. Closure requires a written decision, which may be “corrected source, answer effect inconclusive.” An edit completed today is not a completed measurement cycle today.

Read the source guide: AI Citation Tracking: What to Measure Before You Change Content →

What to preserve: Full answer, observed source links, capture reference and reviewer. Why it matters: Supports the original mention, citation and recommendation decisions.

Read the source guide: AI Citation Tracking: What to Measure Before You Change Content →

What to preserve: Errors, unavailable surfaces and incomplete reviews. Why it matters: Prevents gaps being reconstructed as negative answers.

Read the source guide: AI Citation Tracking: What to Measure Before You Change Content →

What to preserve: Surface, model if known, language, intended market, location evidence, session state and UTC capture time. Why it matters: Makes differences in collection visible.

Read the source guide: AI Citation Tracking: What to Measure Before You Change Content →

What to preserve: Exact wording, IDs, version and frozen file. Why it matters: Prevents a changed question from masquerading as a content effect.

Read the source guide: AI Citation Tracking: What to Measure Before You Change Content →

What to preserve: Page URL, saved content, relevant facts and publication or capture date. Why it matters: Shows what information was available before the edit.

Read the source guide: AI Citation Tracking: What to Measure Before You Change Content →

Sometimes the reader problem requires several coordinated changes. Log all affected pages and describe the intervention as a bundle; the resulting observations cannot isolate each page’s contribution.

Read the source guide: AI Citation Tracking: What to Measure Before You Change Content →

A baseline package makes the old state inspectable. Save the relevant page content and published facts as well as the answer evidence. A screenshot of an attractive result alone cannot show whether a later answer used a changed source. The observation kit contains empty journals and a versioned panel. Its existence does not mean any baseline has been collected. Keep genuine completed observations and potentially sensitive evidence private; publish only the evidence you have permission to share.

Read the source guide: AI Citation Tracking: What to Measure Before You Change Content →

When no before evidence exists, report the completed implementation and its direct checks: the corrected fact is visible, the link works, the metadata matches the page and the deployed URL serves the new content. Keep these checks separate from observed AI outcomes. Collect the first real panel now and label its date honestly. It can become the baseline for the next intervention. The report should distinguish the publication date, the first observation date and any later review date so readers cannot mistake an update timestamp for performance evidence.

Read the source guide: AI Citation Tracking: What to Measure Before You Change Content →

Reuse frozen questions in independent sessions and keep interface, language and collection method comparable. Do not reuse a successful answer under a new record ID. A new model, a changed location setting or a revised prompt can be worth studying, but belongs in a distinct comparison group. OpenAI documents that memory can personalize responses and searches. Record the session conditions you can observe rather than assuming that a fresh chat removed all personal context. Preserve unknown settings as unknown. Choose an observation window that permits repeated collection and record when source access was checked. There is no universal waiting period that proves an assistant has incorporated an edit. A before-and-after difference remains descriptive because other sources, systems and buyer context may also have changed.

Read the source guide: AI Citation Tracking: What to Measure Before You Change Content →