Citation share

How to measure it

Fix a question set — typically 50 to 200 questions your buyers actually ask. Run them on a schedule across each engine, record which domains are cited, and compute your share of total citations. The absolute number matters less than its trend and your share relative to named competitors.

What moves it

In practice three things dominate: whether a page contains a clean answer block for the exact question, whether the author entity resolves, and whether the claim appears anywhere else on the open web to corroborate it.

Building the question set

The measurement is only as good as the questions. Build the set from three sources: the questions buyers ask in sales calls, the queries that already drive branded and non-branded search, and the comparison questions people ask before choosing a vendor. Freeze the list so the metric stays comparable across runs, and version it when you add questions rather than editing in place.

• —Definitional: 'what is an incentive audit?'
• —Procedural: 'how do you redesign a compensation incentive?'
• —Comparative: 'incentive design vs. change management'
• —Commercial: 'who does cultural performance audits?'
• —Entity: 'who is Aaron Bare?'

A simple scoring model

For each question and engine, record cited (1) or not (0), plus whether the description of you was accurate. Citation share is total citations divided by total question-engine pairs. Track accuracy separately — being cited and misdescribed is a different problem from not being cited at all, and it is usually an entity problem rather than a content problem.

Leading indicators when the sample is small

Citation share moves slowly and noisily on a 50-question set. Faster signals: whether a newly published answer block appears verbatim in any answer within two weeks, whether your named framework starts being used as a generic term, and whether referral sessions from ChatGPT, Perplexity, and Copilot appear in analytics at all.

Frequently asked

Is citation share the same as impressions?

No. Impressions count how often a link was shown. Citation share counts how often a source was used to construct an answer, whether or not anyone clicked.

What is a good citation share?

There is no universal benchmark. Judge it against your own baseline and against the specific domains that currently occupy your question set.