Ask the Question Yourself
Procurement teams already put supplier questions to ChatGPT and Perplexity the way they once put them to Google. Ask about olive oil, tomato paste, or bottled water from Iraq or Kurdistan. The answers are thin: regional directories, Turkish firms with strong websites, listings last updated years ago. This is not the model disliking the region. There is simply nothing local for it to cite. Filling that empty seat is the job of generative engine optimization: structuring your company so the answer names you.
Why the Seat Is Empty
An AI engine builds answers from crawlable text and consistent facts: a company name, location, and product list stated the same way across a site, its schema, and the directories that mention it. The regional standard of an Instagram page, a WhatsApp number, and a PDF price list is invisible to that process. A firm without machine-readable facts does not rank lower. It does not exist in the answer set at all.
First Movers Buy the Answer Cheap
In crowded markets, being the cited supplier is expensive. Here the channel is empty. A few structured pages and consistent entity facts can make one firm the answer for an entire category. And answers have inertia: engines keep citing what they have already cited. Claiming the empty seat now costs a few pages of work. Unseating whoever claims it will cost years.
If I rank on Google, will ChatGPT recommend my company?
No. Google rankings and AI assistant answers are separate contests, and winning one does not carry to the other. Ahrefs measured the overlap between Google's top results and the sources AI assistants cite: 12% on average, with Perplexity the outlier at 28.6% and the other assistants near 8%. An exporter on page one of Google can still be absent from every assistant a buyer asks.
The engines also disagree with each other. Muck Rack's citation analysis found Claude cites Reuters 50 times less often than ChatGPT and leans on a different set of publications entirely. Each assistant keeps its own list of trusted sources. The empty seat is not one seat. It is one per engine, per question, per language.
For a supplier in Iraq or Kurdistan this cuts both ways. A Google ranking will not put you in the answer. But you do not need one to get there. A crawlable page that states verifiable facts can be cited by an assistant before it ranks anywhere.
How does a company in Iraq get cited by ChatGPT and Perplexity?
By publishing the facts an engine needs in a form it can read: a crawlable site in English and Arabic, entity schema, and pages that state what you supply, at what capacity, from where. The engine's job is assembling a confident answer from verifiable text. Your job is being the easiest verifiable text in the category.
The checklist is short. One company name, address, and product list, stated identically on the site, in Organization and Service schema, and in every directory that mentions you. A page per product category that answers the buyer's question in its first sentence. A named person with a work email, because an answer that ends in a contact gets acted on. Then corroboration: chamber listings, trade portals, mentions on crawlable third-party sites, so the model can confirm you exist outside your own domain.
This is the program we run as generative engine optimization services. None of it is exotic. It is entity hygiene applied to a market where almost nobody has done it.
Which buyer questions should my company be the answer to?
The ones procurement types: who supplies your product from Iraq, who manufactures it in Kurdistan, which firms export it to the buyer's market. Write that list before you write a single page. A short list covers most categories: the product plus Iraq, the product plus Kurdistan, the product plus export, and the same questions in Arabic.
Then ask each question yourself, in ChatGPT and Perplexity, in English and Arabic. Record which firms get named and which questions come back with directories and guesswork. The questions that return guesswork are your open seats. Each one becomes a page that answers it in the first sentence, in the buyer's language.
Arabic matters more than most owners expect. Gulf buyers ask in Arabic, and the pool of crawlable Arabic sources on Iraqi suppliers is close to zero, so one native Arabic page faces almost no competition for the citation. This is the ground our SEO in Iraq program covers: the audit, the pages, and the monthly re-run of the same questions to prove the seat is yours.
