Stop selling to humans.Win the AI that answers them instead.
The loop
When a buyer asks AI what to buy, it names a handful of brands and skips the rest. Korpora measures where you land in those answers, across every model and language, the gap between what you spend and what AI actually says, and how to close it. Start free with your AI visibility score.
Free, no call: we measure where AI recommends you across every model and hand you the gaps. What we measure · See case study.
External-only: we never touch your systems, and we don't retain or sell your data. How we handle data.
The loop
The shift
AI is the new front door.
The buyer's first question, “what's the best option,” now goes to ChatGPT, Perplexity and Gemini before it ever reaches Google. The model compares you against thirty competitors in seconds and names a handful before a human ever sees your page. Everyone else is invisible.
The brands AI names get the buyer. The rest never enter the room, no matter what they spent to get there.
This is the new top of funnel, and unlike every channel before it, you cannot buy your way in.
Ask · this page
the buyer asks AI, not Google
Name
it returns a handful of brands
Buy
those few get the click and the sale
Desktop search and mobile each took five to ten years to mature. This shift is adopting in twelve to eighteen months.
The problem
If AI doesn't name you, you don't exist.
Ask the model for the best in your category and you are usually not in the answer. Not because you are worse, but because the AI builds its recommendation from the sources it trusts, reviews, Reddit, YouTube, editorial, and your brand is not in them. Your ad spend and your homepage never enter the calculation.
Your ads never count
The model can't see a frame of your Meta or TikTok spend. Walled gardens, no transcript, invisible to AI.
It trusts third parties
AI weights independent sources over your own site. Of course your homepage says you are the best. It wants corroboration.
Absence is invisible
If you are not in the reviews, threads and articles the model reads, you are not a candidate. You never see the deal you lost.
Google rank · ad spend · AI recall, three different scoreboards
18.5%
Eight Sleep runs the most Meta ads in its category. AI names it less than 1 in 5.
Korpora measurement
31%
Ranking #1 on Google is only a 31% chance of landing in the AI answer.
NP Digital · 4,300 prompts
75%
Of AI citations go to sources that never appear in Google's top 10.
NP Digital
AI visibility
Win the sources AI pulls from.
The loop, not the score
You do not win the answer on your own website. The models build their recommendation from the sources they trust. Here is where AI is actually pulling your category's answer from, and how to own those surfaces:
Third-party presence
Show up favourably in the reviews, Reddit threads and listicles the models actually quote. This is the corpus the answer is built from.
Video and transcripts
Ungated YouTube with clean, indexed transcripts. The open channel the models feast on, while your TikTok and Meta spend stays invisible.
Editorial and Wikipedia
The independent authority AI leans on for “best X.” Earned coverage and accurate reference pages move the answer.
Owned content it can quote
Comparison pages, an llms.txt and structured pages the model can lift and cite directly, in metrics, not adjectives.
Start with the comparison page. It converts 6.8% of AI traffic, the highest of any page type, ahead of your homepage and far ahead of a blog post (NP Digital). It is the first asset we ship.
Who it's for
Where the AI answer decides the sale.
Any brand a buyer would ask AI about is already being named or skipped, whether they track it or not. A few of the places that is already deciding who wins:
Consumer & ecommerce
“Best mattress for back pain?”
The AI names three brands and the buyer never opens a tab to compare. If you are not one of the three, the sale happens without you. We measure where you land and the sources that put the other three there.
Considered purchase
“Which tool should we use for this?”
Buyers ask AI to shortlist software, vendors and services before they ever reach your site. The shortlist is the whole game. We show you whether you are on it, against whom, and why.
Global & multi-model
Every model, every market.
The answer differs across ChatGPT, Gemini, Perplexity and the Chinese models, and across languages. We measure each one, so you see where you win, where you vanish, and which market is leaving you out.
Agencies
Sell the answer before clients ask the question.
Every brand you run is being ranked by ChatGPT and the Chinese models, named or skipped. Scan any client in seconds, show them where they place against rivals, and add AI visibility to the retainer before it becomes table stakes.
Not listed? If a buyer could ask AI about you, you can be measured. Get your free scan.
Get your free
AI visibility
score
We measure where AI recommends you across every model, who it names instead, and the sources feeding the answer. Your scorecard in 24 hours. Free, no sales call.