New · one-click Verify re-checks your live site and stamps fixes verifiedNew · one-click Verify for your fixes →Now tracking Google AI Overviews alongside ChatGPT, Claude, Gemini & PerplexityNow tracking Google AI Overviews →Free AI visibility audit: see where you're invisible in ~15 minutesFree AI visibility audit in ~15 min →New · one-click Verify re-checks your live site and stamps fixes verifiedNew · one-click Verify for your fixes →
AI visibility platform

Become the brand
AI recommends.

Track how often ChatGPT Claude Gemini Perplexity Copilot ChatGPT recommends you. Optimize. Get cited. Grow everywhere AI answers.

Track visibility
Actionable fixes
Win everywhere

We check ChatGPT, Claude, Gemini & Perplexity for your business, free report in a few minutes.

✓ No credit card✓ Ready in ~15 min✓ 4 engines checked
Measured every day across ChatGPT Claude Gemini Perplexity Copilot
On · running· example market view, 90 days scanned daily
Visibility Score82+34 in 90 days
AI ranking · your marketscore
#1On82+3.4
#2Nike80-0.4
#3Adidas76+0.8
#4Hoka69+1.4
#5New Balance62-1.2
#6Asics55+0.2
Crawlers+12%
1,240AI crawls this week
Live answers4 engines
best running shoes for marathon trainingnow
on cloudmonster vs nike pegasus8s ago
best daily trainer for beginners12s ago
cushioned shoes for long runs17s ago
Next actionssorted by impact
+6.1Own the 'daily trainer' comparison queries›
+4.2Add Product schema to the Cloudmonster page›
+3.7Get cited in 'best running shoes' roundups›
[01] The shift / why now

One answer. Two or three names. No page two.

AI search is new, but how it works isn't a mystery. Here's what happens between a customer's question and a booking.

2–3names

That's the whole shortlist an AI gives back.

When a customer asks who to choose, the engine compiles its trusted sources in seconds and answers with two or three names. You're one of them, or you don't exist for that customer.

observed across our daily scans, every engine
ChatGPT ⌄
Ask anything↑

Simulated conversation — run the free audit to see your real answers.

aCustomers ask AI, not Google

Instead of scanning results, people ask ChatGPT, Gemini or Perplexity directly — and get one answer back, not a list to browse.

bThe AI shortlists 2–3 businesses

The engine compiles trusted sources in seconds and names a few businesses. Everyone else is invisible; there is no page two.

cIt chooses on signals, not quality

Structured data, quotable answers, machine-readable reviews, presence on sources AI trusts. All buildable — most businesses just haven't.

dGet in early and it sticks

Engines re-check and re-learn constantly. Build your signals now and you keep the recommendation as AI search grows.

[02] Measure / evidence, not a mood ring

Numbers you can defend, from answers you can read.

Four measurements, taken from the real answers your buyers get. Scroll through what each one shows you.

01
Visibility Score

One number for how often AI names and recommends you — measured across every engine you track, every day.

from a real answer · “best running shoe brand right now?”

…the names that come up most often include On, Nike and Hoka. On's Cloudmonster line keeps being recommended for daily training…

Onnamed ✓recommended ✓position #1
Every answer scored → one daily number82/100
02
Prompt watchlist

Every buyer question, engine by engine: recommended, mentioned or absent — and how each one moves over time.

best running shoes for marathon training✓✓○✓
daily trainer with good cushioning✓○✓—
on cloudmonster vs nike pegasus✓✓✓○
beste hardloopschoenen voor beginners○—✓○
✓ recommended○ mentioned— absent
03
Citations

The sources AI actually reads when it talks about your market, and which of your own pages earn a cite.

Where AI reads about this market
Reviews 44% Editorial 27% Community 19% Other 10%
Top cited sources
runnersworld.com 21 reddit.com 14 g2.com 9 on-running.com 8
04
Share of voice

Your slice of AI answers versus the competitors the engines actually name — not the ones you guessed.

Share of AI answers naming each brand
On34%
Nike28%
Hoka17%
Others21%
How these numbers are made

[03] Fix / the part nobody else does

Most tools stop at the score. We do the work.

Every scan turns into prioritized fix cards grounded in your real site, never generic advice. Then we apply what can be applied, and verify the rest.

01
Grounded fix cards

Each card says why it matters for your visibility, exactly how to do it on your platform, and ships a ready prompt or a generated file.

02
Applied for you

Schema, llms.txt and robots served on your live site. WordPress, Astro and Next.js connect natively; articles publish on approval.

03
Verified, not assumed

One click re-checks your live site and stamps a fix verified, or tells you exactly what's still missing.

found by your scanYour site has no llms.txtAI crawlers get no guide to your pages · high impact
Citeon generates the file
applied to your live site WordPressAstroNext.js
✓ verified live
79/100named in 2 more answers this week
One fix, start to finish. Your dashboard runs this for every card.
[04] Prove / only tool that shows the money
See the customers AI sends you.
Not just the mentions.
One tiny cookieless line on your site counts the visitors ChatGPT, Perplexity, Gemini and Copilot send you — and the calls, emails, WhatsApp messages, bookings and forms they turn into. No Google Analytics needed, no developer, nothing to configure. Live in about two minutes.
Others reading GA4
Depend on your Google Analytics. Leads only count if you already configured GA4 key events — most local businesses never have.
Others going server-side
Need setup inside your CDN or server logs. Great for counting AI crawlers — but they don't track a single lead.
Citeon
One line, cookieless, works on any site. Visits and leads from AI, attributed per engine, from day one.
  • Calls
  • Emails
  • WhatsApp
  • Bookings
  • Forms
Numbers you can trust are numbers explained. Read exactly how we measure →
[05] Proof / measured, same rubric, no articles published
An Italian restaurant in The Hague, before and after.
Six fixes over one quarter — the largest a canonical tag pointing their whole site at a dead domain. Measured by our own daily scans.
MetricBeforeAfter
Named in AI answersshare of answers naming them, across ChatGPT, Claude and Perplexity4.7%→57%
Visibility Scorementions and recommendations combined, 0–1005→55
Site readiness for AIour GEO audit, re-run on the same rubric45→80
Method: the same prompt set, engines and rubric before and after. Surfaces we didn't measure are reported as gaps, never as zeros. The numbers plateaued when the work paused — which is exactly why this is a subscription.
[06] Pricing / cancel any time
Start with visibility. Upgrade when you want it fixed for you.
Yearly saves 15% for the first 20 accounts. Monthly starts with a 3-day free trial.
Visibility€89 /monthFor businesses that want to see where they stand — and what to fix.
  • 50 prompts, organized by buyer intent, tracked daily
  • Your 3 chosen AI engines
  • Fix cards with ready prompts + generated files
  • Citations, competitors, sentiment
  • Unlimited team members
Start tracking
Optimize done for you€299 /monthFor businesses that want the work done, measured and proven.
  • Everything in Visibility
  • All engines + Google AI Overviews, daily
  • Files applied for you, full GEO audit + monthly report
  • Content autopilot, 10 optimizations a month
  • Up to 3 projects, one shared prompt pool
Get started
EnterpriseCustom pricingFor brands and agencies that need more than 3 projects.
  • Custom prompts, projects & engines
  • Dedicated onboarding & support
  • White-label reporting, on request
  • Everything in Optimize, at your scale
Contact us
Your customers already ask AI. 
Make sure it says your name.
Track every engine daily, get the fixes, and watch the mentions come to you.
Get started
Not ready? Run the free audit first →