How to Read the GEO Dashboard: Citation Rate vs. Share of Voice
Use Content Hub MCP to track AI visibility, and distinguish citation rate from share of voice before interpreting any numbers.
What is AI visibility?
AI visibility, also called Generative Engine Optimization (GEO), measures how often and how prominently your brand appears when users ask AI models such as ChatGPT, Claude, Gemini, and Perplexity questions relevant to your space. An example is "best developer marketing agency for AI startups." As discovery moves from search queries to chat interfaces, a model's decision to cite you, and where it places you among its sources, plays the role that organic ranking plays in traditional SEO.
Content Hub tracks AI visibility by running a defined set of prompts against multiple AI models on a recurring basis. It records which sources are cited, how often, and in what position. When Claude Desktop is connected to Content Hub MCP, you can retrieve this data in a chat.
Citation rate vs. share of voice
The two metrics measure different things. A high value on one and a low value on the other is not a contradiction. It is a specific, diagnosable pattern.
| Citation Rate | Share of Voice |
|---|---|
| How often you are cited, across every prompt run against every AI model. | The percentage of all citations in a cluster, including competitors, that go to you. |
| Answers: "When someone asks a relevant question, does an AI model mention me?" | Answers: "Among everyone who is mentioned, how large is my share?" |
| Expressed as the percentage of evaluations that included you, for example "cited in 464 of 985 evaluations." | Expressed as the percentage of total citation volume across all competing domains, for example "7% of all citations in this cluster." |
| A single-brand presence metric. | A relative, competitive dominance metric. |
A brand has a high citation rate and a low share of voice when it is mentioned often, but always alongside several other sources. The walkthrough below shows this pattern.
Before you start, confirm the following:
- A GEO cluster exists with prompts added to it.
- The cluster has completed several GEO runs, so citation and ranking data is available.
Fetch the GEO cluster overview
Request the full data for the Developer Marketing GEO cluster in the Infrasity organization.
Claude locates the cluster and returns a summary:
- The number of tracked prompts.
- The number of prompts with zero citations.
- The number of prompts that recently moved up or down in rank.
- The competitor domains that appear most often, including domains not formally tracked as competitors on the cluster.
Claude then groups the 40 tracked prompts into three categories:
- Strongest performers: rank #1 with a high citation rate.
- Rank declines: prompts that lost position, each named with the competitor that now ranks higher.
- Zero-citation gaps: mostly pricing, ROI, and geographic-scope questions. Infrasity already wins the "best agency" queries.
Review share of voice
Request the share of voice for the cluster.
Claude retrieves the pre-aggregated dashboard data instead of tallying across prompts, and returns a ranked domain table. Infrasity leads with 7% share of voice, 2 points ahead of the nearest competitor.
Claude also identifies two factors that qualify the lead:
- The citation rate behind the lead has fallen sharply in recent days.
- When Infrasity is cited, its average rank in the answer is #13.
Infrasity leads on breadth, not on prominence or recency.
Review citation rate
Request the citation rate for Infrasity in this cluster.
Claude reuses the data from the dashboard call. Infrasity is cited in 47% of evaluations (464 of 985) over the last 30 days.
Claude then shows the daily trend behind the average. The rate peaks at 57% in late July and falls to 0% within a couple of weeks, and sample sizes shrink toward the end of the period. Claude offers two possible explanations: fewer prompt runs were scheduled in that window, or citations dropped. Investigate either cause.
Compare both metrics
Request a comparison of citation rate and share of voice, with insights and takeaways.
Claude returns a side-by-side table and resolves the apparent contradiction. Infrasity is cited in nearly half of all evaluations but holds only 7% share of voice. Most answers cite several sources, sometimes 8 to 10 domains, so Infrasity is mentioned often without dominating any single answer. This matches the average rank of #13: broad presence with limited authority.
Claude concludes with these takeaways:
- The strength is breadth, not depth.
- The citation rate decline is the most urgent issue, because it also threatens the share of voice lead.
- Improving rank position is the highest-leverage action, because Infrasity is already cited often.
- Closing even a few zero-citation gaps raises both metrics.
This workflow takes a single cluster overview to a metric comparison with prioritized takeaways in one Claude Desktop conversation.
Sign in to app.infrasity.com and review your own GEO dashboard to compare your citation rate and share of voice.
