Rankings tell you where a page sits in Google. They don’t tell you whether ChatGPT recommends you, whether Gemini names a competitor instead, or whether Google’s AI Overview cites your product page or someone else’s.
AI share of voice answers that question. It’s the metric I rely on most when running GEO (generative engine optimization) and AEO (answer engine optimization) programs, and you can start measuring it today with nothing more than a spreadsheet.
Key takeaways
- AI share of voice is the percentage of brand mentions in AI-generated answers that belong to you, compared with the competitors you track.
- The formula is your brand’s mentions ÷ total mentions of all tracked brands × 100. Report it per platform as well as blended.
- Test a focused set of about five prompts across Google AI Overviews, ChatGPT, Gemini and Perplexity, under consistent conditions.
- Track mentions, citations and links separately. Citations and links are what send people to your site.
- Label branded prompts and run each prompt two or three times, because AI answers vary from run to run.
What AI share of voice measures
Quick answer
What is AI share of voice?
AI share of voice is the percentage of brand mentions in AI-generated answers that belong to your brand, measured against the competitors you track. It shows how often tools like ChatGPT, Gemini, Perplexity and Google's AI Overviews name you when people ask about your category.
If an AI answer about design tools names five brands and you’re one of them, you hold part of that answer. Across a set of prompts and platforms, those mentions add up to a share you can track over time.
It sits alongside two related signals, and it’s worth recording all three separately:
- Mention: The AI names your brand or product in its answer.
- Citation: The AI lists one of your pages as a source.
- Link: The answer links directly to the page you’re trying to promote.
A mention builds awareness. A citation and a link are what send people to your site.
The formula
Quick answer
How is AI share of voice calculated?
Divide your brand's mentions by the total mentions of all tracked brands across your prompt set, then multiply by 100. For example, 16 mentions out of 51 total gives an AI share of voice of about 31%.
AI share of voice = (your brand’s mentions ÷ total mentions of all tracked brands) × 100
“All tracked brands” means you plus the competitors you’ve chosen to watch. Pick them once and keep the list stable, or your numbers won’t be comparable month to month.
A worked example
Say you market a design app and you’re tracking four competitors. You run five prompts on three AI platforms and count every brand mention in the answers.
| Platform | Your mentions | All brand mentions | AI share of voice |
|---|---|---|---|
| Google AI Overviews | 8 | 20 | 40% |
| ChatGPT | 5 | 16 | 31% |
| Gemini | 3 | 15 | 20% |
| All platforms | 16 | 51 | 31% |
The blended number, 31%, is useful for trend reports. The per-platform numbers are where the decisions are. In this example you’re strong in AI Overviews and weak in Gemini, so Gemini is where the next round of work should focus.
How to run the test, step by step
Quick answer
How do I measure my brand's share of voice in AI search?
Run a focused set of customer-style prompts in each AI platform under consistent conditions, and record every brand each answer mentions. Calculate your share per platform and overall, then repeat monthly to track the trend.
1. Build a focused prompt set
Start with the questions your content answers, the questions customers actually ask, and the follow-up questions AI tools generate on their own (often called query fan-outs). In my testing, a tight set of five prompts produced more relevant and consistent results than ten, so start small and go deep.
2. Choose your platforms
Track Google AI Overviews, ChatGPT, Gemini and Perplexity at minimum. They disagree with each other more often than you’d expect: I’ve seen the same prompt earn strong brand visibility in ChatGPT and none at all in Gemini.
3. Keep conditions consistent
Use a fresh chat for every prompt, test in a private or logged-out window where you can, and run each round from the same location within the same week. AI answers shift with context, so consistency is what makes your numbers comparable.
4. Record more than mentions
Quick answer
How can I tell if ChatGPT mentions my brand?
Ask ChatGPT the questions your customers ask, each in a fresh chat, and note whether your brand is named, cited as a source or linked. Repeat each prompt two or three times, since answers vary, and add a follow-up asking for recommended tools.
For every answer, log whether your brand was mentioned, cited and linked, its position in the answer, which competitors appeared, and the overall sentiment. A spreadsheet with one tab per platform keeps this manageable. It’s the same structure as the tracker I built for an enterprise GEO program.
5. Ask the follow-up question
Many first answers explain a topic without recommending any tools. Adding a follow-up such as “Include the top tools to use” often surfaces brand mentions that weren’t in the first response. Record follow-up results in their own column so you can report both.
6. Track over time
Re-run the same prompts monthly, and after every major content launch. AI answers drift as models and sources change, and the trend line is what tells you whether your work is moving the needle.
Mistakes that skew the number
- Counting branded prompts as wins. If a prompt includes your brand name, the AI will almost always mention you. Track branded prompts separately, or label them clearly in your reports.
- Treating one run as a result. The same prompt can return different answers minutes apart. Run each prompt two or three times and average the results.
- Blending mentions and links. A mention without a link still matters, but it doesn’t drive traffic the way a citation does. Report them separately.
- Reporting only the blended number. A healthy overall share can hide a platform where you’re invisible.
Tools that help
Quick answer
What tools track brand visibility in AI answers?
AI-visibility tools such as Semrush's AI visibility toolkit, Ahrefs Brand Radar and HubSpot's AEO Grader track brand mentions and citations across AI platforms. A spreadsheet works well for manual tracking, and a quick manual check confirms any tool is reading the right signals.
Manual tracking in a spreadsheet works well, especially at the start, and tools add speed and scale as your prompt set grows. Whichever you use, spot-check its results by running a few prompts yourself. For Google specifically, Search Console now includes an AI performance report showing how your pages perform in AI Overviews and AI Mode.
What to do with your number
If your share is low on a platform, look at the sources that platform cites in its answers. Those are the pages shaping the AI’s view of your category, and getting your brand featured accurately on them is often the fastest way in. Keep your brand name close to the topics you want to own, in your own content and in listings and articles elsewhere. Then measure again.
Frequently asked questions
What's a good AI share of voice?
There's no universal benchmark, because it depends on how many competitors you track and how crowded your category is. Compare your share with your top competitor's, and focus on the trend: a share that rises month over month means your work is landing.
Why does my brand show up in ChatGPT but not in Gemini?
Each AI platform draws on different sources and weighs them differently, so visibility in one doesn't guarantee visibility in another. Check which sources each platform cites for your prompts, and focus your efforts on the sources the weaker platform relies on.
Can I improve my AI share of voice?
Yes. Get your brand featured accurately on the pages AI platforms cite, publish clear content that answers your customers' questions directly, and keep your brand name close to the topics you want to own. Then measure again each month to see what's working.
Summary
AI share of voice shows how often AI assistants name your brand compared with your competitors, and it’s simple to measure: run a focused set of prompts across the major AI platforms, count every brand mention, and divide yours by the total. Report it per platform, keep mentions, citations and links separate, label branded prompts, and repeat the test monthly. The trend tells you whether your GEO work is paying off, and the per-platform gaps tell you where to focus next.
If you’d like to know where your brand stands in AI answers today, let’s talk. An AI visibility audit is often the best first step.