What version of your brand is AI showing customers today?
I help companies see themselves and their competitors as AI does.
Only through measuring AI responses can companies manage their brands across AI sources.
I measure brands in AI systems and recommend changes to improve those results over time.
You’ve earned it. Now make sure AI knows it.
Seeing is believing. Enter this prompt into your chatbot:
"What are the best [your subcategory] for [your target audience]?"
If you don’t like the answer, let’s talk.
Is this SEO for AI?
No, a brand’s own sites are only 5–10% of the sources Google’s AI Overviews draw on.¹ SEO is important, but it comes later.
Google’s AI Overviews now reach more than 2.5 billion people a month.²
Only 16% of Fortune 500 consumer brands systematically track AI search performance.¹
Tracking is only the start. I also research the sources that inform AI about you and your competition, whether they are accurate (often they are not), and how to fix problems.
¹ McKinsey & Company, “New front door to the internet: Winning in the age of AI search,” October 2025. The 16% is from a survey of about 30 Fortune 500 consumer brand CMOs, September 2025.
² Google I/O keynote, 19 May 2026.
Why measure brands in AI sources?
“Optimization without measurement is guesswork.
A brand cannot know whether its GEO investment is working, whether its share of voice in AI responses is growing or shrinking, or whether a competitor is gaining ground unless there is a common standard for measuring any of it.”
IAB “Measuring Visibility in the AI Era”, August 2026
My method for measuring brands in AI aligns with the IAB’s guidelines. It is also informed by 25+ years of brand marketing services, client partnerships, and first-hand understanding of how businesses grow.
HOW THE MEASUREMENT CYCLE WORKS
01
Start with “category questions”.
Ask the questions a customer asks before they know your name: to learn, to compare, for a recommendation, and to buy.
Measure
02
Many prompts, many answers
Three AI systems, eight times per question. Google AI Overviews, three times by hand. Each answer saved word for word and dated.
03
Count names and sources cited
Which brands each answer includes, how often, plus every web page it cites as a source for its response.
04
Open every cited page
The brands’ own pages are set aside: the method measures what others say about you.
Verify
05
Find the pages that define the category
Directories, rankings, and lists that name several competitors. These shape who AI recommends.
06
Plot every brand
How often AI names each brand when asked about the category, against how many of the category-defining pages carry each brand name.
07
Compare with your facts
Your Brand Ground Truth: what your site says and what you want to say, checked against what AI says.
Act
08
Prioritize what can move the needle
Rank the gaps that can improve your visibility. Find good evidence to support each recommended fix.
09
Measure again
Same questions every quarter to see which corrections AI noticed. Monthly, to monitor changes in the category, sources, and technology.
Much of the counting is automated.
Researching and mapping what to repair - that’s the work.
FAQ
“Don’t we have this already?”
-
Unverified numbers from an AI are no way to make decisions, per the IAB.
I verify the numbers. I also verify the sources and show you which ones carry you and which ones left you out.
-
No. They answer different questions. A tool watches the number. I explain it.
-
That is good. But what moved it? A source you control, a competitor’s error, or a change in how the system answers can all move the score. The score alone can’t tell you.
-
To know how AI represents your brand depends on opening each cited page and checking what it actually says.
Software that only counts and lists AI output can misrepresent your reality.
“How is this different from SEO?”
-
No, setting up measurement of your brand across AI sources should come before optimizing the sources you can control.
“Optimization without measurement is guesswork. A brand cannot know whether its GEO investment is working, whether its share of voice in AI responses is growing or shrinking, or whether a competitor is gaining ground unless there is a common standard for measuring any of it.”
IAB “Measuring Visibility in the AI Era.”
-
I ask category questions a customer would ask, without the brand’s name in them. We agree on the questions together.
Asking by name shows what AI knows about you. It doesn’t show whether AI recommends you - or even associates you with your category.
-
AI does, mostly. The starting roster comes from my asking AI who competes with you, which you review and edit.
Then the category answers themselves add brands - names neither of us put on the list. Sometimes they aren’t the ones you would have picked, and that is itself a finding.
-
Yes. You can be named often and also be described incorrectly or confused with another organization.
That is why I measure what AI says about you, not just whether it names you.
“What do we get, and why does it need you?”
-
Every answer is read by two different AI models, and any disagreement is decided by me. Every cited source is opened. Every number is dated. Transparency and traceability are core to my process.
Each measurement cycle uses the same frozen questions, so the before-and-after comparison holds. My process follows the IAB’s 2026 “Measuring Visibility in the AI Era.”
-
There are a few reasons:
Independence: It is generally better if the team doing the work is not also reporting what worked. However, when a company lacks a marketing or communications capability, I can assist with implementing recommended changes.
Disruption: Making changes to AI sources can be disruptive to the organization. For example, one recommendation might be to reimagine your website so that AI can better associate your brand with the category of products and services you offer. Your product marketing strategy, production workflow, agency ecosystem, business priorities: all can be involved at different times to implement those changes. It is better to have an internal working group communicate the goals and manage the process internally.
-
An AI visibility measurement report that makes the disclosures the IAB asks of measurement providers in its August 2026 framework, “Measuring Visibility in the AI Era,” and states where it meets the decision-grade standard.
Your Brand Ground Truth. A table of facts I have sourced from your public information and the evidence I have surfaced. Together, they define your brand in AI systems.
A ranked list of fixes, each with the evidence to back it up.
Optional: a board presentation and a working session with your team.
-
Changes to some AI sources can take a long time to be produced and seen by AI systems. For example: updates to Wikipedia, your website, pitches to influential industry publications, and local press.
-
Yes - technology, competitors, and news can change overnight. The AI chatbots customers use change without notice, a competitor improves its profile in an influential source, a news article describes the brands in your category - all reasons to look more often than quarterly.
-
AI tools are used throughout the measurement cycle, but the first cycle for one brand in one category takes about 120 hours of human time, and each quarterly reading, about 40.
-
This work is for whoever has to coordinate the in-house team and the agencies around one picture.
Often, an in-person working-session/workshop to understand the process, goals, timelines and to communicate those to the departments within the client’s organization that may be asked for (public) information or changes.
Though information in sources about brands can be static for years, the number and nature of those sources is constantly changing.
The category, too, is in motion: competitors, partners, and publications gain and lose influence over time.
AI technology and commerce platforms used by consumers can change overnight.
It is why measuring your brand and competitors in AI systems is so important.
How We Help
Brand AI Measurement
Your brand is in pieces across AI sources. Which pieces, which sources, and from when - these decide how AI describes you. Measurement shows you the gaps between the brand you control and the sources you don’t. I prioritize those gaps by their impact on your brand’s visibility to consumers, and recommend fixes for your staff and agency partners to execute.
Brand Ground Truth
A table of every major claim your brand makes that confirms its existence and the products or services you provide. Each fact is mapped to what can be established in public sources: registers, accreditors, trade bodies, parent brands. Where the proof is wrong, stale, or missing, you see the evidence gaps and the sources that need attention. The Brand Ground Truth can become a central resource for every communication that describes your brand.
Timeline: 6 Weeks
Quarterly or Monthly
Measurement Cycles
Sources change, models change, competition changes, and your business changes. Each quarter (or monthly), I re-run the audit to measure and report how these changes are impacting AI’s synthesized portrait of the brand.
Workshop
I also offer a half-day workshop, in person, to start the project. Your working group - the people who will run the process internally and report to stakeholders - learns how the measurement works, what the sources are, and what their role in the process is each quarter. After the workshop, your team executes; I measure and advise.
Executive or
Board Presentation
Reporting the findings of a measurement cycle to a board is a different job. I did the measurement, so I can explain it - the process, the findings, and the recommended course of action - and answer the questions a dashboard can’t. So I also offer to present: top-line for boards and senior executives, or a more detailed version for knowledge-transfer to your staff and agency partners.
Michael Quinn developed our method and tools, and he runs the audits. He has 20+ years of experience building agencies for luxury, hospitality, and tech brands in New York, San Francisco, Paris, and Hong Kong. Member, International Association for Safe & Ethical AI.
Betsy Pearce has spent 20 years advising top executives and creative leaders across fashion, luxury, and beauty in Europe and the United States. She leads client relationships.