The New KPIs: Mention Rate, Citation Rate & Share of Voice in AI Search
"What position do we rank?" is starting to feel like asking how many fax numbers you have. When roughly 93% of AI Mode sessions end without a click, your rank in a list nobody scrolls to matters far less than whether the AI named you in its answer. That shift demands a new scorecard - and most teams are still reporting on the old one.
Quick answer: The four KPIs that replace keyword rank in AI search are mention rate (how often you're named), citation rate (how often your URLs are sourced), share of voice (your presence vs competitors), and sentiment/accuracy (how you're described). Measure them by running a fixed set of buyer prompts across engines on a schedule and counting the results.
Why don't keyword rankings work anymore?
Keyword rankings measure your position in a list of links - but AI answers often replace that list entirely. With around 93% of AI Mode sessions ending without a click, a user may read a synthesized answer and never see the ranked results at all. If your brand isn't in the answer, your rank is irrelevant to that session. The metric didn't get worse; the surface it measured disappeared.
What is mention rate vs citation rate?
Mention rate is how often an AI engine names your brand across your tracked prompts; citation rate is how often it actually links to or sources your specific URLs. They're related but distinct: you can be mentioned without being cited (named, no link) or cited without a strong mention. Tracking both tells you whether you're known (mention) and whether your content is doing the work (citation).
| KPI | What it measures |
|---|---|
| Mention rate | % of tracked prompts where your brand is named |
| Citation rate | % of answers that link to or source your URLs |
| Share of voice | Your mentions vs competitors on the same prompts |
| Sentiment | Whether mentions are positive, neutral, or negative |
| Accuracy | Whether the AI describes you correctly |
What counts as a good share of voice?
A good share of voice is leading your direct competitors on the prompts that matter to your buyers - not dominating the whole web. Since even the most-cited domain on a platform rarely exceeds 5% of all citations, absolute numbers look small by design. Measure share of voice inside your category, against named rivals, on your priority prompt set.
Key Insight
Accuracy is the KPI everyone forgets - and the most dangerous to ignore. An AI that confidently mentions you with the wrong pricing, features, or positioning is actively misinforming buyers at scale. A high mention rate built on wrong facts is a liability, not a win.
How to measure them
The method is the same whether you do it in a spreadsheet or a platform: fix the inputs, then count consistently. Define a stable prompt set, run it on a schedule across engines, and record each KPI the same way every time so your trend lines are comparable week to week.
- 1 Fix your prompt set: Choose the buyer questions you'll track and don't change them casually.
- 2 Run on a schedule: Weekly across ChatGPT, Perplexity, Gemini, and AI Overviews.
- 3 Count consistently: Mention rate, citation rate, share of voice, sentiment, accuracy - same rules each run.
- 4 Benchmark competitors: Track the same prompts for rivals to make share of voice real.
- 5 Act on accuracy first: Fix wrong AI descriptions before chasing more mentions.
"Around 93% of AI Mode sessions end without a click - so being named in the answer matters more than ranking in a list nobody scrolls to."
-- AI search analytics, 2026
Frequently asked questions
What is the difference between mention rate and citation rate?
Mention rate is how often an AI engine names your brand across a set of prompts. Citation rate is how often it links to or sources your specific URLs. You can be mentioned without being cited, or cited without a prominent mention, so tracking both gives a fuller picture of AI visibility.
What is a good share of voice in AI search?
There's no universal benchmark because it's relative to your competitors and category. A useful target is leading your direct rivals on your priority prompts. Since even the most-cited domain on a platform rarely exceeds 5% of all citations, share of voice is best measured within your niche, not against the whole web.
How are AI search KPIs different from keyword rankings?
Keyword rankings measure where your page sits in a list of blue links. AI KPIs measure whether you appear inside the generated answer itself. With around 93% of AI Mode sessions ending without a click, being named in the answer matters more than ranking in results a user never scrolls to.
How do I measure these KPIs?
Run a fixed set of buyer prompts across the major engines on a schedule, then count: how often you're named (mention rate), how often your URLs are sourced (citation rate), your share versus competitors (share of voice), and the tone and correctness of mentions (sentiment and accuracy). A spreadsheet works to start; tools automate it at scale.
Figures in this article come from third-party industry research published in 2025-2026. We summarize and link the originals below; numbers are directional findings from those studies, not guarantees.