What is mention share?
Mention share is a metric that shows, for AI answers to a specific topic or question, each brand's count of answers that mention it as a share of the sum of that count across all tracked brands. Because it is each brand's share among the tracked brands, the mention shares of all tracked brands add up to 100%. Like share of voice (SOV) in advertising and media coverage, it expresses a share within the set of brands being compared.
In traditional SEO, the main metrics were search rankings and organic traffic. In the AI search era, however, more and more users finish gathering information from AI answers before they ever visit a website. That is why a new metric is needed to measure how much presence your brand has in AI answers.
That metric is mention share.
How to calculate mention share
Basic calculation steps
Mention share is calculated as follows:
- Step 1: Send the AI several prompts related to a specific keyword (e.g., "good sleep")
- Step 2: Analyze the AI's answers and detect mentions of each brand
- Step 3: Divide each brand's count of answers that mention it by the sum of that count across all tracked brands
As a formula:
Mention share (%) = number of answers that mention the brand ÷ sum of that count across all tracked brands × 100
Calculation example
Here is a concrete example. Suppose you send ChatGPT five prompts about the keyword "good sleep" and get the following results:
- Answer to prompt 1: mentions SleepWell, NightRest, and DreamPillow
- Answer to prompt 2: mentions SleepWell and NightRest
- Answer to prompt 3: mentions NightRest, DreamPillow, and SleepWell
- Answer to prompt 4: mentions SleepWell
- Answer to prompt 5: mentions NightRest and DreamPillow
In this case, the number of answers that mention each brand is:
- SleepWell: 4 (mentioned in 4 of the 5 answers)
- NightRest: 4 (mentioned in 4 of the 5 answers)
- DreamPillow: 3 (mentioned in 3 of the 5 answers)
The sum of the three brands' counts is 11 (4 + 4 + 3), so their mention shares are:
- SleepWell: 36.4% (4÷11)
- NightRest: 36.4% (4÷11)
- DreamPillow: 27.3% (3÷11)
Calculating across multiple AI platforms
For a more accurate picture of mention share, it is important to run the same analysis on multiple AI platforms such as ChatGPT, Gemini, and Claude and look at the results as a whole.
Each AI platform tends to answer differently, so looking at only one AI can lead to skewed conclusions. As explained in How brands get recommended in AI search, each AI has its own characteristics.
Why mention share matters
1. A metric that reflects how search behavior has changed
In traditional search, users looked at a list of results, clicked a link, and visited a site. With AI search, however, more and more users make decisions just by reading the AI's answer.
In other words, even if you rank first on Google, if your brand does not appear in the AI's answer, you are missing the chance to be seen by AI search users.
2. It quantifies a brand's presence in AI answers
"Is our brand visible in AI search?" Mention share lets you answer this vague question with a quantitative metric. Once it is quantified, you can:
- Report and explain results to management more easily
- Measure the impact of initiatives (how mention share changed before and after them)
- Compare your brand with competitors more clearly
- Set goals with it as a KPI
3. It can be tracked over time
By measuring mention share regularly, you can track how it changes over time. This lets you:
- See the impact of generative engine optimization (GEO) initiatives as a trend
- Spot competitors' moves early
- Detect the impact of changes to AI algorithms
For example, if your mention share rises 5 points after a major media outlet publishes a comparison article featuring your brand one month, you can test the hypothesis that media coverage increases your exposure in AI search.
How mention share differs from similar metrics
Mention share vs. search rankings
Search rankings show where a specific web page appears in the results, while mention share shows a brand's presence across AI answers as a whole. It is not unusual for a brand to have high search rankings but low mention share.
Mention share vs. mention rate
A metric that is easily confused with mention share is mention rate. Both count the number of answers that mention the brand, but they divide by a different number (the denominator), so they give different values even when calculated from the same answers.
- Mention rate = number of answers that mention the brand ÷ number of valid answers (all answers except those that could not be retrieved). Because it is calculated for each brand separately, the rates do not add up to 100%.
- Mention share = number of answers that mention the brand ÷ sum of that count across all tracked brands. Because it is each brand's share among the tracked brands, the shares add up to 100%.
For example, suppose brand A appears in 5 of 10 valid answers and brand B appears in 3. A's mention rate is 50% (5÷10), and B's mention rate is 30% (3÷10). If A and B are the only tracked brands, A's mention share is 62.5% (5÷8), and B's mention share is 37.5% (3÷8).
Because mention share is relative to the other tracked brands, it rises when competitors appear less often, even if your own brand's count does not change. To see how often your brand itself appears in AI answers, check its mention rate as well.
Mention share vs. brand awareness
Brand awareness is the share of consumers who "know" a brand, while mention share is how often AI answers name the brand (its share among the tracked brands). Even if awareness is high, whether AI recommends the brand is a separate question.
How to use mention share in GEO
1. Measure your baseline
Start by measuring the current mention share of your brand and your competitors. Once you register keywords along with your own and your competitors' brands, AIOPulse automatically measures mention share in three AIs (ChatGPT, Gemini, and Claude) every day, using a fixed set of Japanese-language prompts.
2. Set goals
Once you know your current numbers, set specific goals:
- Raise mention share from X% to Y% within three months
- Narrow the mention share gap with key competitor A by Z points
- Reach a mention share of X% or higher in a new keyword area
3. Run initiatives and measure their impact
Carry out content and PR initiatives, and monitor changes in mention share regularly. Run a PDCA (plan-do-check-act) cycle: strengthen the initiatives that work and rethink those that have little effect.
4. Analyze by LLM
Analyzing mention share by AI platform shows you which AIs your brand is strong in and which it is weak in. For example:
- High mention share in ChatGPT but low in Gemini → strengthen your visibility in Google Search
- Few mentions in AIs that use web search to answer (Gemini and Claude in AIOPulse's measurements) → publish more up-to-date content that search can easily reference
Limitations and caveats of mention share
Mention share is a powerful metric, but it has some limitations:
- AI answers can differ each time: the same prompt can return different answers depending on timing and context. You need a large enough sample to get statistically significant results
- The quality of mentions needs to be evaluated separately: what matters is not only whether the name appears, but also whether the brand is recommended in a positive context
- Its correlation with direct business results is still being established: research on how much mention share affects conversions is still at an early stage
Summary: A new KPI for the AI era
Mention share is one of the most important metrics for measuring brand presence as AI search spreads. You could call it the core KPI of the AI search era, the counterpart of search rankings in the SEO era.
Thinking about the ROI of GEO explains the relationship between mention share and business results in more detail.
If you want to know how often your brand appears in AI answers, contact us. AIOPulse measures mention share in ChatGPT, Gemini, and Claude every day, based on their answers to questions asked in Japanese.