What you measure

AI share of voice

See your share of the brand mentions in AI answers, next to every competitor, across the prompts that matter in your category.

Start trackingBook a demo

Free first scan. No card needed.

Up to seven models · Refreshed daily · Every share links to the answer behind it

Nike / Dashboard
Share of voice32.2%0.5pp
50%25%0%
26 Sep5 Oct15 Oct25 Oct
Nike's daily share of voice held between 31.1% and 33.3% from 26 Sep to 25 Oct 2026.
Brands by share of voice
  1. 01NikeYou32.2%0.5pp
  2. 02Adidas28.1%0.5pp
  3. 03New Balance18.4%0.4pp
  4. 04Puma14.2%0.7pp
  5. 05Under Armour7.0%0.2pp

Example brand shown

Tracked across up to seven models
  • ChatGPT
  • Claude
  • Google AI Overviews
  • Google AI Mode
  • Microsoft Copilot
  • Perplexity
  • Gemini

What does AI share of voice measure?

AI share of voice is your brand's percentage of all the brand mentions in AI answers to your prompts, measured against each competitor you track.

We measure it across a set of prompts, not one question. One answer tells you very little, because the same prompt can name different brands on different days. Ask hundreds of prompts every day and the shares settle into a number you can compare: you hold 32% of the mentions in your category, your nearest competitor holds 28%, and the gap is closing or growing.

  • 01

    Your share

    Your share of all the brand mentions, next to each competitor's.

  • 02

    Per model and market

    The same share for each model and market, so you can see where a competitor is winning.

    All models →
  • 03

    Against each competitor

    Every competitor's share on the same scale, so the gap between you is one number.

  • 04

    Over time

    Your share day by day, so you see a competitor gaining ground early.

Visibility and citations are measured separately: see AI visibility tracking and AI citation tracking.

How is AI share of voice calculated?

Share of voice is your brand’s mentions divided by the mentions of every brand you track. We count them across all the answers to your prompts, for a set period, model and market.

300mentions of your brand
divided by
900mentions of every tracked brand · 1,000 answers
equals
33%your share of voice

ExampleAcross 1,000 answers to your prompt set, the models name a tracked brand 900 times. If 300 of those mentions are yours, your share of voice is 33%. Your competitors' shares make up the other 67%, so the tracked field always adds up to 100%.

How we measure share of voice →

Why do rankings not tell you this?

An AI model returns one answer, not ten positions. You are named in it or you are not.

A results page has ten positions, so rank tracking can tell you that you’re third and a competitor is fifth. An AI model gives one answer. There’s no position one to hold. It writes a short reply, names a few brands, and stops. You’re either named or you’re not, and the answer can change each time the prompt is asked.

So the comparison that matters is how often you appear against how often they do, on the same prompts and the same days. Your rankings can stay steady while a competitor becomes the default recommendation for your category. Rank tracking still does its job on the results page. It just can’t see inside the answer, and that’s where the recommendation happens now.

Compared onRank trackingAI share of voice You
Looks atThe results pageThe answer itself
MeasuresWhere a page sits among ten positionsHow often you're named next to each competitor
OutputA positionYour share of brand mentions, over time
CompetitorsPages ranking around youBrands named instead of you
EvidenceThe results pageThe answer behind it

How is AI share of voice tracked?

Your prompts and competitorsNike – GB
Prompts · 50
  • Basketball10
  • Football boots10
  • Lifestyle trainers10
  • Running10
  • Training & gym10
Competitors · 5
  • Adidas
  • New Balance
  • Puma
  • Under Armour
  • Gymshark
Step 01 · Define

Define your prompt set and competitor set.

Start with the questions customers ask in your category: research, comparison and buying prompts, for each of your markets. Then pick the competitors you want to be measured against. You approve both lists, so every share is based on prompts you recognise and rivals you really compete with.

Share of voice by modelYouAdidas
  • Google AI Mode34.4%
  • Gemini32.4%
  • Google AI Overviews32.1%
  • ChatGPT30.4%
Step 02 · Ask

We ask every prompt on up to seven models, every day.

Each prompt runs on ChatGPT, Claude, Google AI Overviews, Google AI Mode, Microsoft Copilot, Perplexity and Gemini, every day, and we save every answer. Same prompts, same day, so you can compare shares between models and from week to week.

Share of voice by topicNon-brand prompts
TopicNikeAdidasNew BalancePuma
Training & gym39.2%31.6%3.1%5.2%
Basketball27.0%25.1%14.6%18.6%
Football boots30.1%29.9%15.2%24.3%
Lifestyle trainers27.0%31.0%29.7%12.0%
Running30.6%26.0%30.7%12.4%
Step 03 · Compare

See your share next to every competitor.

We note which brands each answer names. Those add up to your share of the category's mentions, by model, by market and over time, next to each competitor's. Every number links back to the answers behind it, so you can back it up when someone asks.

What you can do with it

Spot movement

Catch a competitor gaining ground

If a rival's share climbs 4 points in a month, you'll see it weeks before it shows up in your pipeline. You can see which models and prompts are driving it, and which pages the models use when they name them.

Prove movement

Show leadership a number that moves

“We went from 22% to 31% of category mentions this quarter” is something a board understands. Every point links back to real answers, so you can explain how it was measured.

Find absences

Find prompts where you are missing

Some prompts never name you, on any model. Those are your clearest gaps: questions your customers ask where a competitor is the only answer. We list them so you can work through them.

Built for teams reporting upward.

AI share of voice is the competitor line in your monthly report: one number, against named rivals, moving over time. SEO teams use it to show leadership how they compare in AI search, the way rankings once did for the results page. Agencies run it for every client in its own workspace, and report each client's share against that client's competitors.

AI share of voice, answered.

How is AI share of voice calculated?

Your brand's mentions divided by all tracked-brand mentions across the answers to your prompt set, expressed as a percentage. If the models produce 1,000 answers and your brand is named in 300 of the 900 brand mentions, your share is 33%. It is calculated per model, per market and overall.

Which AI models do you track?

Up to seven, depending on your plan: ChatGPT, Claude, Google AI Overviews, Google AI Mode, Microsoft Copilot, Perplexity and Gemini. Every model gets the same prompts on the same day, so you can compare them fairly and any gap between models is real.

How many competitors can I track?

You choose a named competitor set per brand, and you can change it as your category shifts. Share of voice is measured across the brands you track, so your share and your competitors' shares add up to the whole tracked field, per model, per market and overall.

How often is the data refreshed?

Daily. We re-ask all your prompts on every model on your plan every day, and each day's answers update your share and your competitors' shares. That's often enough to catch real movement early, and steady enough to report on month to month.

Know who the models name instead.

Connect your brand, approve your prompts and competitors, and your share of voice builds daily from your first scan.

Start trackingBook a demo

Up to seven models · Refreshed daily · Every share links to the answer behind it