AI Citation Tracking: Measure AI Citations and Visibility

What is AI citation tracking?

AI citation tracking is the practice of checking, on a repeatable schedule, whether AI answer engines such as ChatGPT, Perplexity, and Google AI Overviews cite or name your website when they answer the questions your buyers ask. You run the same fixed set of prompts on each engine, record which sources each answer cites, and compare the results over time.

AI citation metrics it involves:

Citation rate: C / P: the share of eligible prompts where the engine cites your URL or domain.
Mention rate: The share of eligible prompts where your brand is named in the answer, with or without a link.
Citation target: Whether the answer cites a specific URL, your domain, your brand name only, a competitor, or no source.
Competitor citations: Which other sources the engine cites for the same prompts.
Scope: The engine, prompt-set version, and observation window every number above depends on.

Google Search Console impressions and clicks are not AI citation metrics; they report classic web results only.

How to measure AI citations

Method status: This is NRLC’s defined method. It is not an industry standard, and it is not a published result.

To measure AI citations, run a frozen prompt list against each engine, record whether the answer cites your URL, your domain, your brand name only, a competitor, or no source, then compute citation rate and mention rate for that window. Google Search Console clicks and impressions are not citation rate.

Citation rate: NRLC defines citation rate as the share of eligible prompts in a fixed prompt set where a named AI engine returns a citation or source attribution to the monitored URL, domain, or entity during a stated observation window.

Under the NRLC measurement convention, citation rate = C / P. C is the number of eligible prompts with a qualifying citation; P is the total number of eligible prompts evaluated. Every reported rate must identify the engine, prompt set, denominator, and observation window.

Frequently Asked Questions

What is AI citation tracking?

What is citation rate and how is it calculated in AI search monitoring?

NRLC defines citation rate as the share of eligible prompts in a fixed prompt set where a named AI engine returns a citation or source attribution to the monitored URL, domain, or entity during a stated observation window. Under the NRLC measurement convention, citation rate = C / P. C is the number of eligible prompts with a qualifying citation; P is the total number of eligible prompts evaluated. Every reported rate must identify the engine, prompt set, denominator, and observation window.

How do I measure AI citations?

What is the difference between mention rate and citation rate?

NRLC defines mention rate as the share of eligible prompts where the monitored entity is named in the answer, whether or not a URL is cited. A mention without source attribution is not counted as a citation. Citation rate requires source attribution to a URL or domain.

Does Google Search Console measure AI citations?

No. Search Console reports classic web results (and, when available, search appearance types such as job listings). It does not report ChatGPT, Perplexity, or Claude citation share. Use it to see which queries Google already associates with a URL, then measure those queries in a prompt set.

How do I track or monitor AI citations and AI visibility?

Track them by repeating the same frozen prompt set on a named engine and comparing citation rate and mention rate only across equivalent observation windows. Monitoring is that repeated measurement. NRLC publishes the method and has not published an observation log, a citation count, or a trend.

How do I know whether AI visibility is improving?

Compare citation rate and mention rate for the same engine, prompt-set version, and eligibility rules across two observation windows. A change is not an improvement until both windows are reported with C, P, and the dates. No such comparison is published here.

How do I measure mentions across AI systems?

NRLC defines mention rate as the share of eligible prompts where the monitored entity is named in the answer, whether or not a URL is cited. A mention without source attribution is not counted as a citation. Measure each engine separately. Do not add engines into one rate unless the prompt set, window, and eligibility rules are stated for each engine.

What cannot be measured from a single AI answer?

A single answer cannot prove ranking, traffic, or that a site is “optimized.” It can only show whether that engine, for that prompt, on that date, cited, mentioned, or ignored a source.

AI Search Measurement definition

What this system owns

Observable: whether a named engine cited a URL, a domain, or a brand name for a named prompt on a named date
Derived: citation rate and mention rate over a frozen prompt set
Not this page: guaranteed Overview slots, “AI traffic,” or Search Console as a citation ledger

Methodology

Freeze the prompt set: Assign the prompt list a version and do not add, remove, or rewrite prompts during the observation window.
Name the environment: Record the engine, interface or mode, account state when relevant, and run date.
Classify each answer: Record a qualifying URL or domain citation, entity mention without citation, competitor citation, or no observed source.
Apply eligibility rules: Exclude failed requests and unusable responses before fixing P, then preserve the exclusion log.
Calculate and report: Calculate C / P and report C, P, engine, prompt-set version, and observation window together.

Measurement limitations

This method measures observed outputs, not private model reasoning. Results can vary by engine, interface, prompt wording, account state, and time. A citation occurrence does not prove that a source caused an answer, produced a click, or generated revenue. NRLC has not published a worked citation-rate example because there is not yet an observation log with a prompt count, engine, dates, citation count, or citation rate.

Related systems

AI SEO — the category this measurement serves
AI Search Optimization — the commercial work that uses this method
AI Search Diagnostics — indexed but not retrieved
AI Search Tools Reality — limits of vendor dashboards
Generative Engine Optimization — retrieval vs ranking

Evidence

Method: NRLC AI Search Measurement method
Observation window: Declared per measurement run; no aggregate run is claimed by these method pages.
Dataset: Versioned prompt set and answer-observation log supplied by the operator
Method: Freeze prompts, name the engine and interface, classify observed citation and mention outcomes, preserve exclusions, then calculate C / P over eligible prompts.
Authors: Joel David Maldonado
Sources: NRLC AI Search Measurement method, Google Search Console Performance report documentation
Last verified: 2026-09-21
Confidence: first-party operational convention; not asserted as an industry standard
Limitations: Observed outputs vary by engine, interface, prompt wording, account state, and time. The method does not expose private model reasoning or establish causal attribution.

View resource Methodology

External scope reference

Performance report (Search results) — Google Search Console Help

AI Citation Tracking: Measure AI Citations and Visibility

What is AI citation tracking?

How to measure AI citations

Frequently Asked Questions

Measurement Guides

Measuring AI Visibility

Tracking AI Citations

Reporting AI Search Performance

Attribution in Zero-Click Search

What Can and Cannot Be Measured

AI Search Measurement definition

What this system owns

Methodology

Measurement limitations

Related systems

External scope reference

Discuss your AI search visibility

Contact