Your buyers are talking to AI. See both sides of the conversation.
We map full customer journeys across ChatGPT, Claude, Perplexity, Gemini and Google AI Overviews: what your buyers ask, what they're told at every stage of the funnel, and the exact turn where you drop out. Tilio is the first UK agency to analyse full customer journeys in AI search, not single prompts.
We'll run a journey from your category before the call and walk you through it live.
[01] The shift
/ where the research now happens
The shift
The most important sales conversation you've never seen
Somewhere right now, a buyer in your category is describing their problem to an AI. They'll ask what their options are. They'll get a shortlist. They'll stress-test it, ask what's wrong with each name, check the practicalities, and decide who to contact. One conversation, start to finish.
No clicks. No session in your analytics. No record it ever happened.
The journey used to be spread across dozens of touchpoints: searches, review sites, comparison guides, forums, your website. Those touchpoints haven’t disappeared. They’ve collapsed into the turns of a single conversation. Google once named that research phase the Zero Moment of Truth. It now happens inside ChatGPT, Claude, Perplexity, Gemini and Google AI Overviews, and it’s often settled before you ever see a click.
Visibility tracking tells you whether you appeared in one answer on one day. It can’t tell you what happens next: whether you make the shortlist two questions later, how you’re described when a buyer pushes back, or whether the facts about you survive to the moment of recommendation. The journey is where deals are won and lost, and until now you couldn’t see it.
[02] A first for the UK
/ what this is, and what it isn't
A first for the UK
A new kind of analysis, and we mean that precisely
Plenty of tools can tell you whether you appear in a single AI answer. We offer that ourselves.
What hasn’t existed in the UK is journey-level analysis: the full conversation, every touchpoint from first question to final recommendation, mapped to the funnel and measured across repeated runs. We built the method and the tooling to do it, and we’ve checked the market: no other UK agency offers it.
A claim like that deserves equal honesty about what this is and isn’t. It’s not a recording of any one customer’s exact conversation, because AI answers vary between sessions. It’s a comprehensive snapshot of the typical journey, built from enough runs that we can show you, with real confidence, what your buyers typically see at every stage.
[03] Specimen journey
/ one conversation, five turns
Specimen journey
Watch a brand disappear
Turn 1
Problem framing
The buyer asks: Our team's outgrown our current setup. What are our options?
The engine answers
The engine explains the category. You're not here yet. Nobody is.
Turn 2
Criteria
The buyer asks: What should we look for in a provider?
The engine answers
The engine sets the buying criteria: flexibility, accreditation, location, support. Sourced from two directories and a competitor's guide.
Turn 3
Shortlist. The moment of truth
The buyer asks: Who are the best providers in the UK?
The engine answers
Four names. Yours isn't one of them. All four appear in the sources from turn 2. This is the turn that decides who gets considered at all.
The shortlist it returns
- Competitor A
- Competitor B
- Competitor C
- Competitor D
Your brand: not on the listTurn 4
Scrutiny
The buyer asks: What about [your brand]?
The engine answers
The engine knows you, but describes a service you retired two years ago and hedges on whether you cover the buyer's region.
Turn 5
Decision
The buyer asks: Which would you go with?
The engine answers
It recommends a competitor. Confidently, with reasons.
The share of journeys where the brand is still in the running, turn by turn. In this composite, the drop happens at the shortlist, before the buyer has heard the brand’s name once.
A composite example. Your analysis is built from your buyers’ real questions and run repeatedly, so findings are frequencies, not one-off anecdotes.
[04] How it works
/ scoping call to executive readout
How it works
Personas, built with your team.
Your sales and BD people already know the questions buyers actually ask. We interview them, then turn that knowledge into buyer personas and the journey scripts to match: opening questions, follow-ups, objections, the lot. We don't invent characters. Personas are rebuilt from the questions your buyers already ask.
Journeys, run properly.
Each persona's journey runs repeatedly across ChatGPT, Claude, Perplexity, Gemini and Google AI Overviews, through all six stages of the funnel: problem framing, category, shortlist, scrutiny, validation and post-purchase. Follow-ups adapt to what each engine says, the way a real buyer's would. Clean sessions, UK conditions, every date and model documented.
Analysis and readout.
Every touchpoint is captured and coded: presence, position, how you're described, which sources were used, where you dropped out and why. Then we present it live to your leadership team, with the priorities ranked.
Timeline: three to four weeks from scoping call to readout.
[05] Scope your analysis
/ live, in about two minutes
Scope your journey analysis in two minutes
Tell us your domain, sector and what you sell, and we’ll draft your buyer personas and the questions they’d actually ask, live, right here. You’ll get a structured brief and an indicative price, and you can bring the whole thing to a scoping call. No payment, no commitment, no forms pretending to be tools.
[06] What we measure
/ the numbers behind the report
What we measure
The metrics behind it
- Share of Journey
- The percentage of journey turns where your brand appears. Share of voice tells you about isolated answers; Share of Journey tells you about the conversation.
- Journey survival curve
- How deep into the journey you stay in the running, and the turn where you drop out.
- Descriptor fidelity
- How the engines describe you versus how you describe yourself.
- Gatekeeper sources
- The small set of third-party sources that decide who makes the shortlist.
- Handoff accuracy
- When you are recommended, whether your services, locations and facts arrive intact.
[07] What you'll find out
/ findings you can act on
What you'll walk away knowing
- What AI actually believes about your brand: what it thinks you do, for whom, where, and at what level.
- Who sets your buyers' criteria: whether the standards buyers judge you by come from you, a competitor, or a directory you've never heard of.
- Exactly why you drop out: unknown, overlooked, disqualified, or fumbled at the handoff. Each one has a different fix.
- What's said under scrutiny: the objections and negatives that surface when a buyer stress-tests you and there's no salesperson in the room.
- Which battlegrounds you can win: where you're strong and weak by funnel stage, persona and engine, and the pipeline at risk in the gaps.
[08] What you get
/ the deliverable set
What you get
- The journey report. The full analysis, written as your buyer's experience, with findings as frequencies across every run.
- Annotated transcripts. Every journey, every turn, every source.
- The journey film. A short cut of screen recordings. Watching your brand vanish from a shortlist lands harder than any chart.
- A live executive readout. We present the findings to your leadership team and take the questions.
- The intervention map. Every drop-off tied to a specific fix, ranked by the pipeline it reopens. Actionable by your team, ours, or both.
Then, if you want it: quarterly re-measurement. The same journeys, re-run, so you can see exactly what moved.
[09] Who it's for
/ and who it isn't
Who it's for
Brands with considered purchases and real competition: where buyers research, compare and shortlist before they ever get in touch.
And honestly, who it isn’t for: if your product is an impulse buy, you don’t need journey analysis. A visibility audit will serve you better.
[10] Proof
/ why the findings hold up
Why Tilio
- The first UK agency to analyse full customer journeys in AI search, not single prompts.
- Delivering with Profound and Noble.
- Working with brands from department stores to scientific research facilities to B2B consultancies.
- A method you can check: every run dated, every model documented, every finding a frequency.
[11] FAQ
/ the questions we get asked
FAQ
What is AI customer journey analysis?
AI customer journey analysis maps the full conversation between buyers and AI answer engines like ChatGPT, Claude, Perplexity, Gemini and Google AI Overviews: the questions asked at every stage of the funnel, the answers given, where a brand appears, how it's described, and where it drops out. Unlike prompt-level visibility tracking, it measures whole conversations, run repeatedly, so findings are frequencies rather than one-off answers. Tilio pioneered the approach in the UK.
How is AI customer journey analysis different from an AI visibility audit?
An AI visibility audit is a snapshot: it tells you whether you're visible in AI search today, prompt by prompt. Journey analysis follows the whole conversation: whether you survive from first question to final recommendation, why you drop out, and what that's costing you. Plenty of clients start with the audit. If you already know visibility's an issue, start here.
Is AI customer journey analysis new?
Yes, and we're specific about what we mean. AI visibility tracking and audits already exist, we offer them ourselves. What's new is journey-level analysis: full multi-turn conversations, mapped to funnel stages, measured as frequencies across five engines. Tilio built the method and the tooling for it, and we keep a dated record of the market to stand behind the claim. If others follow, good: it'll mean the category matters.
Is AI customer journey analysis based on real buyer data?
It's a comprehensive snapshot of the typical customer journey in AI search, not a recording of one individual's exact answers. The journeys are built from your buyers' real questions, and because AI responses vary between sessions, every journey runs repeatedly under controlled, documented conditions. From that data we can give you an extremely confident picture of what a customer typically goes through: what they're told, who they're pointed to, and where you feature. We're precise about what this is and isn't, because that's what makes the findings safe to act on.
Which AI answer engines does journey analysis cover?
Tilio covers ChatGPT, Claude, Perplexity, Gemini and Google AI Overviews as standard. Others, like Copilot or Meta AI, on request.
AI answers change every time. Doesn't that break journey analysis?
It's why the method exists. Every journey runs multiple times per engine, and findings are reported as frequencies: how often you appear, how often you survive to the shortlist. One-off answers are anecdotes. Patterns across runs are evidence.
How long does an AI customer journey analysis take?
Three to four weeks from scoping call to executive readout.
How much does AI customer journey analysis cost?
A Tilio journey analysis is scoped to your category: how many product lines, personas and engines you want covered. Use the builder above for an indicative band, then book a scoping call and you'll leave with a clear fixed price.
What happens after the journey analysis?
You'll have a ranked intervention map. Some clients action it in-house, some hand it to us (content, accuracy work and third-party placements), most do a mix. Quarterly re-measurement is there if you want to track movement against the same journeys.
[12] Next step
/ twenty minutes
Next step
The conversation is happening either way
Your buyers are already asking. The only question is whether you can see the answers.
We’ll run a journey from your category before we speak and walk you through it on the call. Twenty minutes, and you’ll see exactly what your buyers see.