Subscription
Everyone in your category, sat the same exam
Most tools watch your brand. We measure the whole field on one frozen question set across ChatGPT, Gemini, Claude and Perplexity, publish the ranking, and show you what to fix to move up it.
AVI Pro
Everything below, one price
£99/ month
Cancel anytime · billed monthly · no setup fee
What we put to the assistants
- ✓All four engines — ChatGPT, Gemini, Claude and Perplexity. Every one included; none is an add-on.
- ✓15–37 questions real buyers ask in your category, frozen and versioned so scores stay comparable.
- ✓Every competitor in your category, answering the identical set — not a list you have to think up.
- ✓Each question asked repeatedly, because assistants answer differently every time.
What comes back
- ✓How often each assistant mentions you, and how often it actively recommends you.
- ✓A confidence interval on every rate, so you can tell a real move from noise.
- ✓Whether the model already knew you or looked you up — the two need opposite fixes.
- ✓Share of voice against the real field, with one business trading under three names counted once.
- ✓Sentiment, and whether what the assistant says about you is factually right.
- ✓Which sources the assistants cited to answer — including your rivals' own pages.
What you do with it
- ✓A prioritised remediation plan that names the rivals actually taking your share.
- ✓Exact fixes for your site, written for whoever manages it — no jargon, no assumed platform.
- ✓Continuous monitoring with proof-of-impact over time.
- ✓A full technical readiness audit alongside the Outcome score.
What you can show people
- ✓A published, citable position in the index — not a private dashboard nobody else can see.
- ✓An embeddable AVI Score badge for your own site.
- ✓Domain-verified ownership of your brand's entry, so nobody else can claim it.
- ✓Simulated numbers are never published as measurements. If we have not measured it, we say so.
What a reading actually looks like
ChatGPT, Gemini, Claude and Perplexity · 21 July 2026
- Mentioned in
45.6%
of 296 answers judged
How often the assistants brought this brand up at all.
- Really somewhere between
40% – 51.3%
95% confidence
Assistants answer differently every time. Without this range you cannot tell a real improvement from the dice — and it is the number you will spend a budget against.
- Knew the brand already
44%
from training
Earned slowly, through being written about. A long game.
- Looked it up to answer
56%
searched live
Winnable this quarter, by fixing what the assistants can find and cite.
Measured alongside 38 competitors answering the identical questions — which is what turns a number into a position.
Not ready? Run the free scan — it costs nothing and needs no card. Or read how we measure and the questions people ask.