Navigating the AI Search Metrics Landscape

AI search engines don't produce fixed rankings, so the metrics you've relied on for years can't tell you whether your brand is winning. In this lesson, Dillon shows you how to read the signals that matter and make smarter content decisions with confidence.

What You'll Learn

Why the old SEO scoreboard doesn't work for AI search

Traditional SEO measured fixed rankings: position one, position five, page two. AI search engines are non-deterministic. Run the same prompt five times and you'll get different citations each time.

AirOps data shows that less than 10% of the same content gets cited across five consecutive runs of the same prompt.

The question shifts from "where do I rank?" to "how often do we appear, in what context, and with what sentiment?"

This requires a completely different dashboard and a different way of reading your performance data.

Mentions vs. citations: the two foundational Answer Engine Optimization (AEO) metrics

A mention means an AI engine named your brand in its response. A citation means it linked your content as a source. You want both, but they tell you different stories.

Connecting this to the FACT framework:

  • Being mentioned means you've cleared the Findable and Agent-Aligned gates. The model knows you exist and considers you relevant.
  • Being cited means you've also passed the Citable and Trusted gates. The model found your content worth linking as evidence.

Even though AI search is probabilistic, increasing both metrics raises your probability of reappearing in consistent answers.

How to read the mention-citation matrix

Each combination of high/low mention rate and citation rate tells a different story and calls for a different response:

  • High mentions, low citations: Third-party content is carrying your visibility. Others talk about you, but AI engines aren't linking your pages. Focus on creating citable, source-worthy content.
  • High mentions, high citations: You're in a strong position. Defend what you have and expand into adjacent topics.
  • Low mentions, low citations: You're not on the map. Start with gate one of the FACT framework and focus on being findable.
  • Low mentions, high citations: Your content is authoritative on a narrow set of prompts, but you're invisible on the broader landscape. Broaden your topic coverage.

Dillon demonstrates how to use the AirOps platform to filter prompts by mention rate and citation rate (both greater than 60%) to find top performers, and how to catch trending-downward prompts to prioritize content refreshes.

Visibility metrics: mention rate, share of voice, and average position

  • Mention rate is the percentage of AI responses that mention your brand across your tracked prompts. This is your top-of-funnel signal: does the model know you exist when buyers ask questions?
  • Share of voice is your mention rate divided by the total mentions across you and your competitors. Mention rate alone means nothing in isolation. Share of voice tells you whether your number is market-leading or trailing.
  • Average position is where your brand appears in AI responses when mentioned. Position 1 means you're the first brand named. Lower is better.

In the AirOps Analytics section under Visibility, you can see these three metrics over a time series, compare against competitors in a leaderboard view, and break out by topics, platforms, and competitors.

Dillon drills into mention rate by topic, spots a gap on "AI website builder" where competitors are winning, filters by that topic, and breaks down by platform to see which answer engines affect visibility most.

Citation metrics: citation rate, citation share, and the shrinking link economy

Where visibility tells you how often you're mentioned, citations tell you which URLs AI engines are linking to. This space is getting more competitive.

In the last six months, for branded prompts within ChatGPT, the average number of citations per answer fell from five to three. ChatGPT is linking back to fewer sites, which means fewer opportunities for direct referral traffic.

  • Citation rate is the percentage of AI answers that cite your domain out of all answers that include at least one citation. A citation rate of 10% means your content appears as a source in one out of 10 cited answers.
  • Citation share is your citations as a percentage of all citations across your tracked prompts. This is the competitive equivalent of share of voice, but for links instead of mentions.

Reading the AirOps citations dashboard

In the AirOps Citations tab under Analytics, you can see citation rate, citation share, and citation count filtered by topic.

The domain-type breakdown shows you who influences a topic: product sites, competitor domains, media outlets, social platforms, educational sources, and review sites.

As you go deeper, you'll find specific domains like YouTube and Reddit showing up as citation sources.

The top cited URLs table reveals competitors' pages, third-party review sites, and editorial roundups. Dillon highlights that 85% of URLs getting cited in your space are not brand-owned pages.

Visibility and citations: read both, always

AI search visibility has two distinct components:

  • The Visibility dashboard shows whether the model knows you exist and how you compare to competitors.
  • The Citations dashboard shows which content earns the link.

Your habit should be to read both dashboards together. Mentions without citations means others carry your brand story. Citations without mentions means you're authoritative but narrow. The full picture lives in the overlap.

Key takeaways

  1. ChatGPT cut its citations per answer nearly in halfOver the last six months, branded prompts in ChatGPT went from five citations per answer down to three. Fewer links means fewer chances for your content to earn referral traffic from AI search.
  2. 85% of cited URLs in your space aren't yoursThe vast majority of pages getting cited by AI engines belong to competitors, review sites, and editorial roundups. If you're not actively creating citable content, someone else is shaping your brand's AI narrative.
  3. Filter for 60%+ to find your top-performing promptsIn the AirOps platform, filter prompts by mention rate and citation rate both greater than 60%. That intersection shows you exactly which prompts you're winning on, so you can defend them and replicate the pattern.
  4. High mentions with low citations is your content gap, not a winWhen AI engines name your brand but link to someone else's page, you've handed your narrative to third parties. Close the gap by building source-worthy content on your own domain that passes the Citable and Trusted gates of the FACT framework.
  5. Catch downward trends before they become gapsTrending-downward prompts are your early warning system. Spot them in your AEO dashboard and prioritize content refreshes before a competitor claims the territory you're losing.

FAQs

Traditional SEO rankings break down in AI search because AI engines produce different answers on every run. When you test the same prompt five times across platforms like ChatGPT, Gemini, and Perplexity, citation consistency drops below 10%. That means a fixed "rank" is meaningless. Answer Engine Optimization (AEO) requires probabilistic metrics like mention rate and citation rate that measure how often your brand appears across many responses, not where it sits on a single results page.

Resources

  1. How to Measure AI Search Visibility: Step-by-Step Guide for 2026Article
  2. What is Share of Voice?Article
  3. What is Citation Share?Article