Know what AI tells your customers. Then own the answer.
Your customers now ask ChatGPT, Gemini, Perplexity and Google AI before they reach your site. Hey Taylor measures what those engines say about your brand, market by market. We find where they get you wrong or name someone else, and we run the program that fixes it, alongside your team.
Measured across ChatGPT, Gemini, Perplexity, Google AI Mode, Microsoft Copilot and Claude · Every score with its scope declared · A human reviews every report
The answer arrives before your website does.
When a customer asks an AI assistant about your category, your policies or your products, the engine writes the answer itself. It draws on sources you don't control, and it doesn't check with you first. For a large brand that creates three risks.
Wrong
AI states your return window, pricing, eligibility rules or locations incorrectly, and customers act on it. Your support team then corrects what the machine said.
Absent
At the buying moment (“best option for…”, “who should I trust for…”), AI names competitors and directories instead of you, often in markets where you lead.
Out of your hands
AI builds its answer from review sites, forums, listicles and press, not only your own pages. Your site is one voice among many, and often not the loudest.
None of this shows up in your search console or your web analytics. It has to be measured where it happens: inside the answers.
One measure across every market you sell in.
The Trust Index asks the questions your customers actually ask, across six AI engines, and scores what comes back with a versioned, repeatable method. For enterprise brands we run it by market, by product line and by engine, so each team sees its own results.
Findable
Are you named when customers ask about your category?
What your team gets: the share of answers that name you, by market and engine, beside your top three competitors on the same questions.
Accurate
Does AI get your policies, prices, products and locations right?
What your team gets: every misstatement flagged, with the exact wording and the engine that said it.
Recommended
When AI picks, does it pick you?
What your team gets: who AI names instead of you at the buying moment, market by market.
Cited
Where does AI get its answer?
What your team gets: the sources each engine cites, ranked, showing where you appear and where your competitors do.
Engines covered: ChatGPT, Gemini, Perplexity, Google AI Mode, Microsoft Copilot and Claude. Every number carries its scope: the questions asked, the engines, the market and the date.
Your team owns the site. We own the answer outside it.
Most large brands already have strong SEO, content and comms teams. We don't replace them. We add an independent measure and the off-site work that moves AI answers, and we hand your team clear, prioritised fixes for what sits on your own properties.
Measurement
Your team
Owns web analytics and search reporting.
Hey Taylor
Runs the Trust Index across six engines and reports what AI says, by market.
Your website and policies
Your team
Owns pages, product data and policy wording.
Hey Taylor
Flags exactly what AI misreads and why, with fixes ranked by impact.
Press and third-party sources
Your team
Owns brand voice and approvals.
Hey Taylor
Runs the authority program: placements, reviews and references in the sources AI cites.
Legal and compliance
Your team
Approves claims and wording.
Hey Taylor
Works only from approved claims and never publishes unapproved statements.
Leadership reporting
Your team
Takes results to the CMO and the board.
Hey Taylor
Provides a plain-language quarterly read with every number's scope declared.
Measure first. Fix in order. Prove it on the same questions.
Every engagement is scoped to your markets and goals. No public pricing; we quote after the scoping call.
Diagnostic
We scope the markets, product lines and competitors with you, then run the full Trust Index: multi-pass, median-scored, across six engines. You get what AI says about you today, who it names instead, every accuracy flag, the sources it relies on, and a ranked fix list. Your first read arrives within 72 hours of kickoff.
Program
We run the off-site authority work and hand your team the on-site fixes, in order of impact. Accuracy issues on policies and products come first, because customers act on them.
Monitoring
We re-measure on the identical questions, report quarterly, and only call a gain sustained once it has held for three weeks. When an engine changes how it answers, you hear it from us first.
Measured before and after, on identical questions.
B2B IT services and AI-training company
The company needed to own its category's buying questions before its flagship industry event. A press-authority program aimed at the tracked questions took the flagship question from 33–50% visibility to 79% at #1 across five AI engines. The placements now appear as the cited sources inside the answers.
National US market · Direct client
Personal injury firm
Visible on some hometown questions but absent statewide: 7 of 10 tracked questions at 0%. Four months later it held 100% share of voice in its primary market at an average position of 1.3, and statewide questions went from 0% to 22%.
Multi-market, US Southeast · Via partner agency
Every result is published with its scope: the questions, the engines, the market and the dates.
We also measure whether people believe what AI tells them.
The Trust Index started on the street. We interview strangers on camera about how much they trust AI's answers: 100+ unscripted interviews so far, filmed in Toronto, New York and South Korea. In Toronto, the average is 5.6 out of 10. People trust AI for facts they can check and hesitate on judgments they can't. That gap is where your brand's reputation now sits.
The street-interview average is research context. It never appears in any brand's score.
Need to know what AI says about your whole category or region, and whether people believe it? We run commissioned Trust Index studies. Ask about research in the briefing form below.
100+
interviews · Toronto · New York · South Korea
Numbers you can take to the board.
Scope on every number
Each score states the questions asked, the engines, the market and the date. No blended “AI visibility score” without its scope.
A human signs off
Every report is reviewed by a person before it reaches you.
No guarantees we can't keep
The AI engines decide what they say. We promise honest measurement, a sequenced plan and fast response when an engine changes, not a fixed ranking.
Approved claims only
We work from wording your team has approved and never publish statements about your brand without sign-off.
Nothing to install
The diagnostic runs on the public AI engines, the same way your customers ask. We need no access to your systems to start.
Confidential by default
NDA on request, and we complete your vendor onboarding.
The questions in-house teams ask.
We already have an SEO agency. Why add this?
We already have an SEO agency. Why add this?
We already use an AI visibility platform. How is this different?
We already use an AI visibility platform. How is this different?
Can you guarantee AI will recommend us?
Can you guarantee AI will recommend us?
We operate in many markets. How does that work?
We operate in many markets. How does that work?
What do you need from us, and how soon will we see movement?
What do you need from us, and how soon will we see movement?
See what AI tells your customers. Then decide.
A 30-minute briefing. We show you what the engines say about your brand in one market you choose, and how we would scope the full diagnostic. Nothing is committed until you've seen it.
What happens next
1. A 30-minute briefing
We show you a live read on one market you choose.
2. A scoped proposal
Markets, product lines, competitors and timeline, in writing.
3. Your first Trust Index
The full diagnostic, with the ranked fix list. First read within 72 hours of kickoff.