Ship something useful for free.
The public leaderboard and /report/[domain] pages are not trial versions. They are useful on their own. If the free thing is not useful, the paid thing usually is not either.
—— § 01 · About
About the studio that made AIRRNK.
AIRRNK is a product of MATO, a small software studio. We are profitable. We take no VC money. We are not flipping this. What you’re reading right now — the typography, the copy, the decision to ship a free public leaderboard — comes from the same small team that answers every support message.
—— § 02 · The mission
Make the editorial side of AI legible.
Every few decades, the shape of how people find things shifts. Catalogs to directories. Directories to Google. And now, Google to a chat window that answers in paragraphs instead of blue links.
That last shift is already well underway — 47% of under-thirty buyers start their research in ChatGPT, Claude, or Perplexity. The old SEO stack was not built for that. It was built for ranked lists. A paragraph is not a ranked list.
AIRRNK exists because someone has to do the unglamorous work of making a brand legible to a paragraph. To read every sentence a foundation model writes about your category, every six hours, and tell you whether you’re in it.
If we do that well, a lot of the vague anxiety around “AI SEO” becomes boring and fixable. Which is exactly how we like it.
—— § 03 · Why we built it
A short, honest origin story.
“We lost a client because ChatGPT kept recommending their biggest competitor. We ranked #1 on Google. It did not matter.”
In the summer of 2025, one of our consulting clients at MATO — a category-leader B2B SaaS — showed us their pipeline data. Their demo requests were down 40% year over year. Their Google rankings had never been better.
We spent two weeks running queries against every public LLM and Perplexity, manually. The pattern was damning. The model kept naming a competitor with a third of their traffic, because that competitor had a single long-form page that read almost like a textbook. AI loved it.
The tool we wanted did not exist. SEMrush and Ahrefs were busy re-skinning keyword volume. The AI-first tools that existed were thin wrappers around the ChatGPT API, with no proper observation layer, no edge proxy, no way to actually ship a fix.
So we built it. Two months of heads-down work, one very expensive OpenAI bill, and a public leaderboard that went viral on a Tuesday. AIRRNK is what came out the other side.
The client from that first story is now on Scale. Their demo requests are back to baseline. Their competitor is still being recommended — but so are they.
—— § 04 · How we work
Five principles we hold tightly.
Publish numbers, including the bad ones.
Refund rate. Uptime. Median time-to-first-citation. If we show the metric internally, we show it externally. /status and /changelog are the proof.
No VC, no exit plan.
We are funded by MATO’s consulting revenue. We do not take meetings with funds. We do not hire the kind of people whose compensation depends on a valuation event.
Editorial over enterprise.
The interface reads like a small magazine. The documentation reads like a short story. Clean copy is a feature, not a flourish. SaaS does not need to look like every other SaaS.
We answer the email ourselves.
For as long as we can — realistically, until we are a much bigger team — we read every /contact submission ourselves. The day that stops, we will say so.
§ 07 · Start now
See where AI recommends you —
and where it doesn't.
30 seconds. One domain. A full report showing which AI platforms mention your brand, which ones recommend your competitors instead, and exactly how to fix it. Free. No credit card.
Want to track results every week?Create a free account →
One dashboard · six AI platforms
- ChatGPT
- Claude
- Perplexity
- Gemini
- Grok
- Copilot