The verdict in three sentences
A trading house based in Abu Dhabi does almost no product marketing, yet it is now described by AI assistants like ChatGPT to counterparties, trade finance banks, NGOs and recruits who want to know who it is before they talk to it. What determines the quality of that description is factual governance and compliance pages, consistent in English and Arabic, and reliable public mentions by third parties. The subject is managed with a dated record of answers to the same questions, it is priced on quote, and the market charges roughly 6,000 to 15,000 USD for a diagnostic as a 2026 ballpark.
Who asks about a trading house, and what they ask
A trader's audience is narrow and demanding. A credit officer at a bank weighing a financing line, the procurement team of a refinery or smelter, an NGO analyst following supply chains, an experienced trader considering a move to the city: all of them now start with a question to an AI assistant.
| Audience | Question asked (example) | What the AI assistant must be able to find about you |
|---|---|---|
| Trade finance bank | Does this Abu Dhabi trader publish a sanctions compliance policy? | Compliance page, published and dated policy |
| Industrial counterparty | Which trading houses in Abu Dhabi deal in refined oil products? | Precise business line pages, no invented volumes |
| A client's KYC team | Is this company registered in ADGM or onshore? | Legal entity, registration number, address |
| NGO or ESG analyst | Does this trader publish a sustainability or responsible sourcing report? | Report actually published, direct link |
| Trader candidate | Which Abu Dhabi traders are hiring metals operators? | Careers page, offices, culture |
| Trade journalist | Who runs this company and since when? | Leadership page, sober biographies |
| Logistics partner | Does this trader have offices in Singapore or Geneva? | Up-to-date locations page |
A point to watch: legal status. One entity in ADGM, another in a free zone or onshore, a head office elsewhere: if the site does not link these entities clearly, AI assistants mix them up, and a KYC team draws conclusions.
What makes an answer right or wrong
AI assistants like ChatGPT rely on what is public, repeated and consistent. For a trading house, that mostly means solid institutional pages. The rule is simple: publish what is true and verifiable, never invent volumes, counterparties or certifications.
| Page or source | What it must say | What must never appear |
|---|---|---|
| Governance | Board, management, committees, group structure | Vague organisation chart or outdated names |
| Compliance | Sanctions, KYC, anti-bribery, compliance contact | General wording with no document |
| Responsible sourcing | Policies and reports actually published | Unkept commitments or labels not obtained |
| Locations | Entities, addresses, registrations | Closed offices still listed |
| Arabic version | Same content, same date, right-to-left layout | Partial or machine translation |
| Third-party mentions | Trade press, industry associations, registers | Press releases picked up nowhere |
| Structured data | Organisation, entities, contact details | Inconsistent markup across languages |
On personal data, Federal Decree-Law No. 45 of 2021 governs data protection in the UAE, while ADGM and DIFC have their own regimes. The careers page and contact forms must stay consistent with the regime that applies to your entity. This is not legal advice.
What the work covers and what it costs
In broad terms: a review of what AI assistants say about the company in English and Arabic, clarified governance, compliance, locations and careers pages, consistency with registers and public profiles, presence in serious third-party sources in the sector, structured data, then measurement. The aim is not promotional content but making accurate information easy to find. The market is still emerging, and the amounts below are 2026 ballpark figures from typical agency quotes.
| Item | Range (2026 ballpark) | Approximate local equivalent |
|---|---|---|
| Initial diagnostic, English and Arabic | 6,000 to 15,000 USD | 22,000 to 55,000 AED |
| Rewrite of institutional pages | 8,000 to 20,000 USD | 29,000 to 73,000 AED |
| Native Arabic translation and review | 2,000 to 6,000 USD | 7,300 to 22,000 AED |
| Structured data, right-to-left version | 1,500 to 5,000 USD | 5,500 to 18,000 AED |
| Monthly follow-up and dated record | 2,000 to 6,000 USD per month | 7,300 to 22,000 AED per month |
| Trade press relations | 2,000 to 8,000 USD per month | 7,300 to 29,000 AED per month |
At Kolonell the engagement is on quote, depending on your entities, languages and priority audiences.
Measuring: a dated record, not an impression
Need a professional website?
Kolonell builds websites that attract clients, optimized for the Sénégalese market. Free quote in 2 minutes.
The only credible indicator is a dated record of answers to the same questions, in English and Arabic, archived every month. It shows the share of answers mentioning the company, the accuracy of the information (entities, leadership, published policies) and the errors that persist. No serious provider promises an AI recommendation. For a trader, an accurate and sober answer is often worth more than a flattering mention.
Mini case study
Illustrative example: Saadiyat Resources Trading, a metals trader registered in ADGM, tracks 40 questions in two languages, 80 prompts per record. At the start the company appears in 6 answers out of 80, about 7.5%, and several answers still give it a closed office. The working assumption after six months: 20 answers out of 80, 25%, with no location errors. Hypothetical budget: a 9,000 USD diagnostic plus 6 months at 3,500 USD, 30,000 USD in total. If a single senior trader hire comes through a direct application instead of a search firm charging 25% of an assumed 160,000 USD salary, the 40,000 USD saving covers the spend (160,000 x 0.25 = 40,000). These figures illustrate the reasoning, not a result.
FAQ
Does a discreet trading house benefit from AI visibility?
The point is not being visible but being described correctly. A bank or KYC team reading a wrong answer about your entities can delay a relationship by several weeks.
Is an Arabic version essential in Abu Dhabi?
For local audiences, authorities and some candidates, usually yes. A complete native Arabic version with a right-to-left layout costs roughly 2,000 to 6,000 USD in translation on the market.
Can we name counterparties or volumes to be cited more often?
Only what is already public and allowed by your contracts. Inventing or extrapolating volumes is the surest way to have an AI assistant repeat a wrong figure to 100% of the people who ask.
How long before answers are corrected?
As a ballpark, 2 to 6 months depending on how often sources are refreshed. Address and entity corrections often move faster than new mentions.
What does a Kolonell diagnostic cost?
It is on quote, after a conversation about your entities and audiences. As a benchmark, the 2026 market charges roughly 6,000 to 15,000 USD for a bilingual English and Arabic diagnostic.
Find out how AI assistants describe your firm. We record what AI assistants say about your trading house today, in English and Arabic, and list the errors to fix first. Request a diagnostic quote: reply within 48 hours.
Mohamed Bah
Fondateur, Kolonell
Passionate about digital and entrepreneurship in Africa, Mohamed has been helping Sénégalese businesses with their digital transformation since 2020. Founder of Kolonell, he believes every SME deserves a professional and accessible online présence.
