Any brand or entity whose presence in AI-generated search results is being measured by a product.
Marketers or brand managers concerned with visibility in AI search results.
01What it is and how it works
The advertiser is the subject of measurement. You define it by providing a set of keywords, brand names, product names, and related terms. The product then continuously monitors outputs from major AI search platforms—such as ChatGPT, Google Bard, and others—using natural language processing to detect mentions of the advertiser. It records the frequency of mentions, the context (e.g., whether the brand is recommended, compared, or simply listed), the sentiment (positive, neutral, negative), and the prominence within the response (e.g., first mention, last mention, or embedded in a list). The system also tracks competitor mentions to provide share-of-voice metrics. All data is aggregated into a dashboard that shows trends over time.
It's the brand you're tracking in AI search.
02What to do about it
Start by defining your advertiser profile with all relevant brand variations, including common misspellings, acronyms, and product lines. Set up competitor tracking to benchmark your performance. Review weekly reports to identify which queries drive mentions and which do not. If your brand is underrepresented, create authoritative content that AI models are likely to cite—such as well-structured FAQs, original research, and clear product descriptions. Engage with sources that train AI models, like Wikipedia and reputable news outlets. Adjust your strategy based on sentiment: if mentions are neutral, add context that encourages positive framing. If negative, address the underlying issues in your public communications.
03How it is measured or noticed
The product measures several key metrics: mention count (total number of times the advertiser appears in AI responses), share of voice (percentage of mentions compared to a defined set of competitors), sentiment score (aggregated positive, neutral, negative), and position score (where in the response the mention appears). It also tracks the types of queries that trigger mentions—informational, navigational, transactional. The dashboard shows these metrics over customizable time periods, with alerts for significant changes. You can drill down into individual responses to see the exact context of the mention.
04Common mistakes
- Tracking only your exact brand name and ignoring common misspellings, abbreviations, or product nicknames.
- Failing to include competitors in your measurement setup, so you lack share-of-voice context.
- Assuming AI search behaves like traditional search engine results pages (SERPs) and applying the same optimization tactics.
- Not updating your keyword lists when AI models are updated or when new platforms emerge.
- Relying on a single AI platform for your measurement, missing cross-platform differences.
- Overlooking sentiment analysis and focusing only on mention volume.
05Limits and common confusions
The product measures only publicly available AI search outputs. It cannot measure mentions in private, fine-tuned, or enterprise-specific AI models. The data reflects a snapshot of current model behavior; as models are updated, past measurements may not predict future performance. This measurement is often confused with traditional SEO, but AI search optimization differs because AI models generate responses based on training data and context, not just keyword matching. Also, the advertiser is not necessarily the entity that pays for ads—it is simply the brand being tracked. The term 'advertiser' here is used in the sense of a brand that wants to be visible, not necessarily one that runs paid campaigns.
06Worked example
For a coffee brand named 'BrewCo', the product tracked mentions across ChatGPT and Google Bard over a month. It found BrewCo appeared in 12% of coffee-related queries, with 80% positive sentiment. Competitor 'JavaCorp' appeared in 25% of queries. The advertiser used this data to increase content about sustainable sourcing, which improved its share of voice to 18% the next month.
Frequently asked questions
How is an advertiser different from a brand in this product?
An advertiser is the specific entity whose presence in AI search results is measured, while a brand might be a broader concept. In this product, the advertiser is the subject of measurement, and you define it with all relevant variations.
Should I include misspellings and product lines when defining my advertiser?
Yes, you should define your advertiser profile with all relevant brand variations, including common misspellings, acronyms, and product lines. This ensures accurate measurement of mentions in AI-generated search results.
How does the product measure an advertiser's presence?
The product measures mention count (total number of times the advertiser appears in AI responses) and share of voice (percentage of mentions compared to a defined set of competitors or entities). It only measures publicly available AI search outputs.
What happens if I don't define my advertiser correctly?
If you don't define your advertiser correctly, you may miss mentions due to misspellings or variant names, leading to inaccurate share of voice data. This can cause misinformed decisions about brand performance in AI search.
Does the product measure paid ads or only organic AI responses?
The product measures only publicly available AI search outputs, which are typically organic responses. It does not distinguish between paid and organic placements because AI search results are not ad-based in the same way as traditional search engines.
How long does it take to see measurement results after defining my advertiser?
Measurement results appear as soon as the product begins scanning AI search outputs, which is typically within minutes of defining your advertiser profile. However, meaningful trends may require data over several days or weeks.
Can I measure multiple advertisers at the same time?
Yes, you can define multiple advertisers within the product and track each one's mention count and share of voice separately. This allows you to compare different brands or product lines in AI search results.
Asked out loud
spoken, not typedThe same term in the words somebody uses speaking to an assistant rather than typing into a box — written from the situation, which is why each one carries the situation it came from.
It depends on your share of voice compared to competitors. A mention count alone doesn't tell you if you're leading or lagging. Check the percentage of mentions relative to your defined set to see how you stack up.
Update your advertiser profile to include the new product line name and any common variations. The product will then start tracking mentions of that term in AI responses. Do this as soon as possible to avoid missing early data.
First, check that your advertiser profile includes all your brand variations and misspellings. If it's correct, the issue may be that your brand isn't appearing in the AI training data or recent outputs. Consider reviewing your content strategy to improve visibility.