GEO by TriloAI is an AI brand visibility monitoring tool built by TriloAI. It checks whether ChatGPT, Gemini, Perplexity, Google AI Overviews and Claude recommend you or your clients, and turns the results into a report you can deliver, under your own brand if you're an agency.
The questions we actually ask AI
GEO's first case study is our own brand. Every number, every action, and every result below is independently verifiable
Full disclosure: TriloAI is the team that builds GEO. This is not a client case study. We publish it because it's the one case where we can show the entire before-and-after in public.
17 August 2026
| ChatGPT | 5.2% |
|---|---|
| Gemini | 7.9% |
| Perplexity | 0% |
| Google AI Overviews | 0% |
Across 39 cited sources, triloai.org appeared 0 times — the models knew our name, but never retrieved our pages.
All four were implemented — all of them publicly verifiable site settings.
August 2026
Google AI Mode
Listed under professional AI workflow developers in an answer to “which AI automation vendors are there in Taiwan”. That answer cited 10 websites.

ChatGPT
Appears in a vendor comparison table, with the brand name as a clickable citation link — moving from mentioned to cited.

23 August 2026
In our 23 Aug 2026 scan, ChatGPT cited our service page while answering “are there AI automation consultants that specialise in fitness studios and sports venues?”:
https://triloai.org/services/n8n?utm_source=openaiThe utm_source=openai parameter at the end is added by ChatGPT itself — proof it actually retrieved this page, rather than recalling the brand name from memory.
For comparison, in the 17 Aug baseline scan our own domain appeared 0 times across 39 cited sources.
We won't tell you visibility went up X%. This is a single brand, and AI answers vary between runs. We made all four changes at the same time, so we can't tell which one worked, and the major AI platforms haven't publicly committed to reading llms.txt. What we can show is this: the problems the report identified were real, and after fixing them we started appearing where we previously did not.
Result: monthly inquiries from the website form + LINE
530
Before → One month after
Inquiry counts come from the brand's own website form and LINE records.
This is the result of TriloAI's consulting service, not something the GEO software did on its own
The brand asked to stay anonymous, so we don't name it or its locations.
The gym has locations in several countries, yet its website form and LINE together brought in only about 5 inquiries a month. The brand wanted to know what it takes to be recommended when prospective members ask AI for a gym.
What we can't claim: This is one brand over one month. Seasonality, promotions and other factors may also have driven more inquiries, and we can't prove it all came from AI. We didn't run AI-visibility scans before and after, so we don't quote a visibility percentage.
Are you an agency? We don't reach out to brands that use GEO by TriloAI through an agency. Our agency principles →
SEO makes customers find you when they search Google. GEO makes AI recommend you when they ask ChatGPT, Gemini or Perplexity. The entry point to search is shifting from a list of blue links to a single AI answer — and today you need both.
More buyers ask AI, take one answer and decide — without clicking through link by link. They may never see your site, and your organic traffic and inquiries are quietly leaking away.
When someone asks AI to "recommend a product like this," does it name you or a competitor? If you can't see it, rivals are quietly taking customers that were yours — while you have no idea.
Outdated prices, discontinued items, even fabricated claims — AI can present them as fact to your prospects, and you'd have no way to know or correct it.
The terms you keep hearing, in plain words, and how SEO and GEO differ
Buyers now ask AI for the answer instead of clicking 10 blue links. Google ranking still matters, but AI also weighs who gets cited and who is well reviewed
From quick scan to full analysis, one report that serves brand self-audits and agency client work alike
Your company name
AI visibility report
Mention rate
34%
95% interval 22–47%
GEO score
61 /100
Competitor ranking
Lost prompt
“Which gyms in Taipei are good for beginners?”
AI recommended Brand A and Brand B, not you
Next step: publish a beginner class comparison that answers it
Query ChatGPT, Gemini, Perplexity, Google AI Overviews and Claude in one run, with questions in each market's language and local competitors (for the China market: DeepSeek, Kimi and Qwen).
Citability, brand authority, content quality, technical SEO, AI visibility, platform readiness and search presence are weighted into the GEO score, alongside mention rate, ranking, cited sources and lost prompts, so you can see whether AI actually recommends you.
Each report generates content optimization priorities, keyword opportunities, and competitor gap analysis — agencies convert them straight into a client pitch and content roadmap.
PDFs carry your company name, logo and brand colour, in Traditional Chinese, Simplified Chinese, English, Japanese or Korean, ready for cross-border accounts.
After each analysis, the 40 most-mentioned competitors are checked one by one with live web search. Any we can't confirm as a real business are removed, so fewer AI-invented brands slip into your report.
Not a black-box score: every number traces back to a real AI answer
01
We simulate what buyers ask AI across 7 intents: recommendation, comparison, concern, long-tail, local, price and switching. Up to 15 questions, written for your industry and market; the actual count is shown in the report.
02
The same questions go into ChatGPT, Gemini, Perplexity, Google AI Overviews and Claude, recording whether each one mentions you, recommends you and where it ranks you.
03
Mention rates come with a 95% interval; the AI visibility score counts mentions, recommendations and rank; the GEO score weights 7 parts, including citability, brand authority and search presence.
AI answers vary: in our own test the same 14 questions asked twice gave mention rates of 31.5% and 41.7%, which is why every mention rate comes with a 95% interval. See the full method →
Each full analysis uses 3 credits; credits last 12 months and carry over
First, see if AI mentions you
No credit card needed
One shop or one brand
or US$1,980 a year
Running 2–3 brands
or US$3,980 a year
Agencies reporting for many clients
or US$11,980 a year
No account yet? You'll sign up free first, then pay by card on Stripe
Enterprise is quoted on brands, markets and analyses per month. Fill in the form and we reply within 1 business day.
What ChatGPT, Google, Apple and Microsoft changed recently, and what it means for businesses. Every item links to the official source.
Regions not stated
What it means for you: Bing and Copilot now offer official citation data too. Microsoft itself says to treat the numbers as directional signals for planning, not a precise ledger.
Source: Bing Webmaster Blog (opens in a new tab)Taiwan
What it means for you: Buying an ad in ChatGPT and being recommended in its answers are now two separate things. Whether the answers mention you has to be checked on its own.
Source: OpenAI (opens in a new tab)English beta; Japanese and Korean planned
What it means for you: When iPhone users ask Siri directly, brands may show up in the answer. Apple has not said whether answers link to their sources.
Source: Apple Newsroom (opens in a new tab)Worldwide
What it means for you: Businesses can now see their AI Overviews exposure for free in an official Google tool. It covers only Google's own products and counts URL impressions.
Source: Google Search Central Blog (opens in a new tab)The free 1–2 minute quick scan checks how well AI can read your site and runs a live ChatGPT test. The full analysis across 5 AI platforms starts at US$198 a month
Start the free quick scan