Your schema, profiles and bilingual answer-pages, rebuilt so six AI engines can find, trust and repeat your name.
A customer no longer scrolls a page of results and forms an opinion. They ask an assistant, get three or four names back, and pick from those. There is no page two.
It happens inside ChatGPT, Copilot, Gemini, Perplexity, Claude or Grok. No impression, no click, nothing in your analytics. You never learn it happened.
The engine names a handful of businesses. Being the fifth-best-known name in your district is the same as not existing, because the fifth name is not read out.
Reviews, price and photos only matter to someone who already has your name. If you were never named, none of the work you did on those pages gets used.
In this market the same question arrives in Arabic and in English, and the engines answer each one from different source material. A business that is legible in one language and silent in the other is only half in the answer.
We build and monitor both in parallel — same question set, same schedule, reported together.
Everything below comes from our own logged engine runs, with dates. No mock-ups, no borrowed statistics. Largest panel first.
On ChatGPT and Gemini across 9 districts — one query set, logged and dated (single-run screen). Brand size did not protect anyone: well-reviewed, long-established clinics were absent from the answers patients act on. The screening is described in our published method, so you can check how it was run rather than take the number on trust.
Off the shortlist before reviews or price were ever considered.
On a clean, logged-out surface with nothing changed at all — 10 status changes across 36 observations, three windows, July 2026.
ChatGPT, Copilot, Gemini, Perplexity, Claude and Grok. They do not agree with each other. A business cited by one is routinely absent from another, because they draw on different indexes and different source material.
That is also why a single screenshot proves nothing — ours or anyone else's. Client audits use multi-run averages against a fixed question set, and we re-test monthly.
Arabic and English measured in parallel, not translated after the fact.
Legal name Get AI Cited Technology, licensed by Dubai's Department of Economy and Tourism. The number is checkable against Dubai's official business registry.
Method published in full — how we measure · how it works · engine volatility study
"AI Share of Voice" = the share of the questions your customers ask AI where your business gets named — across ChatGPT, Copilot, Gemini, Perplexity, Claude and Grok, in English and Arabic.
One number, reported monthly, whichever way it moves. That is the number we are accountable for.
Nothing here is a black box. Each step produces an artefact you can read, approve or hand to another agency.
We ask the questions your customers actually ask, across six engines — ChatGPT, Copilot, Gemini, Perplexity, Claude and Grok — in English and Arabic, and record who gets named instead of you. You get a one-page snapshot with the questions, the engines, the dates and the names that came back.
You end up with: your starting position, in writing, before you have paid anything.
Structured data, a Google Business Profile rebuild, and bilingual answer-pages — the groundwork that makes your business legible and citable to AI engines.
You end up with: published assets you own, approved by you before they go live.
Your AI Share of Voice in plain numbers, on a simple dashboard — a multi-run average against a fixed question set.
You end up with: a dated record, reported either way, up or down.
Thirty seconds, no form, nothing sent. This is just to tell you whether the free check is worth your time.
If your customers research and you have not seen the answer, you are making decisions about your marketing without knowing what the shortlist looks like. The free check closes that gap — the questions, the six engines, both languages, and the names that came back.
Run my free AI visibility checkIf your customers ask AI before they choose, you need to be in the answer. We started where the stakes are highest.
Dental, aesthetic, dermatology. Our first vertical, with a published screening of 54 Dubai clinics.
Day spas, wellness and med-spa. Our second vertical: engine screening is running now.
Brokerages, legal, business setup: high-ticket, bilingual, competitive — the same playbook ports.
Anywhere buyers ask AI for a recommendation. Start with the free snapshot and see where you stand.
Classic search optimisation is written for a ranking algorithm that returns links. This is written for a model that has to summarise, attribute and name you out loud.
| Classic search work | Answer-engine work (what we do) | |
|---|---|---|
| Goal | Rank a page in a list of links | Get the business named inside the answer |
| Unit of success | Position and clicks | Citations and mentions — AI Share of Voice |
| What the page must do | Match a keyword and earn the click | Answer the question outright, in a form a model can lift and attribute |
| Surfaces measured | One search engine | Six: ChatGPT, Copilot, Gemini, Perplexity, Claude and Grok |
| Language | Often one language, translated afterwards | English and Arabic built and measured in parallel |
| Evidence | Rank-tracker screenshot | Multi-run averages against a fixed question set, dated and logged |
| Stability | Broadly stable week to week | Answers move on their own — we measured 27.8% drift with no intervention |
The two are not in conflict — a well-built site helps both. The difference is what gets measured and what the content is shaped to do.
Not ready for done-for-you? Know exactly where you stand first. Every month: a multi-run report across 6 AI engines — who gets cited for your buyers' questions, who doesn't, and a prioritised fix-list. You implement, we measure. No calls, cancel anytime.
True bilingual coverage — built and monitored in parallel, because your customers ask in both.
We do the work and report results — not another dashboard you have to run yourself.
One clear metric — AI Share of Voice — so you always see what you're paying for.
Everything is publish-ready and approved by you before it goes live.
+ one-off $800 setup and build sprint (free on annual plans). Month-to-month, cancel anytime. We do not guarantee rankings or citations — AI answers change day to day. What we commit to is measurement, reported either way.
Get AI Cited is a Dubai-registered company that helps businesses get named and cited in the answers AI assistants give — ChatGPT, Microsoft Copilot, Google Gemini, Perplexity, Claude and Grok. Its legal name is Get AI Cited Technology and it is licensed by Dubai's Department of Economy and Tourism (DET) under licence number 1635658. The discipline is called generative engine optimisation (GEO), or answer engine optimisation (AEO): structuring a business's public information so that AI systems can find it, trust it, and repeat it when a customer asks for a recommendation. Get AI Cited works in English and Arabic and serves UAE businesses, with a focus on dental, aesthetic and medical clinics in Dubai.
Yes — Get AI Cited is a licensed Dubai business, trading as Get AI Cited Technology under Dubai DET licence number 1635658. It is a professional sole establishment registered in Dubai, United Arab Emirates, issued by the Department of Economy and Tourism. Its permitted activities include web design, internet content provision, social media applications development and management, and blogging services. The licence number can be checked against Dubai's official business registry.
No. SEO is about blue links. This is about being named inside the AI's answer — a different system, increasingly where customers decide.
Your technical build — schema, a rebuilt Google Business Profile and the first bilingual answer-pages — is normally live within the first few weeks. What happens next is not something any agency can honestly guarantee: AI answers are controlled by the engines and change day to day, and citation movement depends on your competitors and your off-site footprint as much as your site. So we do not promise a date or a position. What we commit to is measurement — your AI Share of Voice, measured monthly as a multi-run average against a fixed question set, with the results reported to you either way.
Yes — in English and Arabic, publish-ready, and you approve everything before it goes live.
Because we measured what happens when nothing is changed at all: on a clean, logged-out surface, 27.8% of answers moved by themselves — 10 status changes across 36 observations, three windows, July 2026, published with a DOI. At that level of movement, a promised position would be a sales line, not a fact. What we commit to is a fixed question set, multi-run averages, and the result reported either way.
No. We're a separate, UAE-based, bilingual, done-for-you service.
More detail: full FAQ · measurement method · engine volatility study
Send your business name and website — you'll get a one-page snapshot of your AI visibility, free.