The verdict in three sentences
For a commodity trading house in Singapore, AI assistants like ChatGPT have become a first credibility filter: a credit analyst, a counterparty or a candidate asks them who you are, where you operate and how you handle sanctions and responsible sourcing. A house appears in answers when its public information is consistent, factual and recent, and disappears or is misdescribed when its website comes down to a home page and a logo. The stakes are measured through a dated record of answers to the same questions, and no provider can promise a recommendation.
What your counterparts ask AI assistants
Trading does very little product marketing. Those who query AI assistants are not looking for a random supplier: they are checking. Their questions look like these.
| Who asks | Typical question to an AI assistant | What it checks |
|---|---|---|
| Trade finance bank | "What is known about [name] in Singapore, shareholders and governance?" | Structure, management, track record |
| Industrial counterparty | "Which metal ore traders have a Singapore office and a supplier due diligence policy?" | Presence, published policy |
| Compliance officer | "Has [name] been mentioned in a sanctions case?" | Reputation, no ambiguity |
| NGO or ESG analyst | "Which Singapore-based agricultural traders publish a sustainability report?" | Reports actually published |
| Trader or operations candidate | "Working at [name] in Singapore, culture and prospects?" | Recruiting, team, values |
| Insurer or charterer | "Which oil products trading companies operate from Singapore?" | Activity, chartered fleet, offices |
| Business journalist | "Main players in thermal coal trading in Southeast Asia?" | Market position, spokespeople |
The concrete risk is less absence than error: an AI assistant that confuses your company with a namesake, cites a former executive or links your name to a case involving a competitor. For a trading house, such confusion can slow a credit line or an account opening.
Why some houses appear and others do not
AI assistants rely on what is published and repeated consistently. Well-described houses share a few traits: factual pages on governance and offices, a compliance policy actually published (sanctions, know your customer, responsible sourcing), dated ESG reports where they exist, third-party mentions (specialist press, professional associations, public programmes such as Enterprise Singapore's Global Trader Programme when the company holds and announces that status), and structured data that describes the organisation without ambiguity.
Conversely, a minimal site, entity names that differ across registries, undated PDF reports or empty careers pages leave AI assistants to fill gaps with third-party sources, sometimes old ones. Nothing should be invented: no volumes, no counterparties, no certifications. For entities that fall under the MAS for a financial activity, public communication follows its own rules. Contact and application forms fall under the PDPA, overseen by the PDPC. This is not legal advice.
What the work covers and what it costs
In broad terms, the work covers a review of what AI assistants say about the company today, alignment of public information (site, registries, profiles, press releases), factual pages on governance, compliance and recruiting, and tracking over time. It often ties in with existing financial communication and media relations.
| Market service | Local range (SGD, 2026 ballpark) | Approximate equivalent (USD) | Note |
|---|---|---|---|
| Initial diagnostic | 3,000 to 10,000 | 2,300 to 7,700 | Depending on number of entities and businesses |
| Monthly follow-up | 2,500 to 8,000 | 1,900 to 6,200 | Record, adjustments, report |
| Writing compliance and governance pages | 1,000 to 3,000 per page | 770 to 2,300 per page | Legal review extra |
| Specialist media relations | 4,000 to 12,000 a month | 3,100 to 9,200 a month | Financial communication agencies |
| Structured data and technical clean-up | 2,000 to 6,000 one-off | 1,500 to 4,600 | Often tied to a redesign |
These figures come from typical agency quotes and public rate pages, converted at roughly 1 SGD to 0.77 USD. The market is young and gaps are wide. Kolonell works on quote.
How to judge results
Honest measurement rests on a fixed list of 20 to 40 questions, asked at regular intervals to AI assistants like ChatGPT, with each answer dated and archived. Three things are tracked: the share of answers that cite the company, the accuracy of what is said (entities, offices, executives, activities) and whether the right pages appear as sources. A persistent error on a sensitive fact, such as confusion with a sanctioned namesake, becomes a correction priority.
Need a professional website?
Kolonell builds websites that attract clients, optimized for the Sénégalese market. Free quote in 2 minutes.
Mini case study
Illustrative example: Lumen Straits Commodities, a 60-person agricultural products trader in Singapore, finds that AI assistants misdescribe its offices and ignore its sourcing policy. Management commissions a USD 5,000 diagnostic and then USD 3,000 a month of follow-up for 6 months: 5,000 + 6 × 3,000 = USD 23,000.
Market assumption for comparison: a recruitment firm often charges 20 to 25% of annual salary. For a trader hired at USD 180,000, 25% comes to 180,000 × 0.25 = USD 45,000. If a better-described careers page enables a single direct hire, the spend is covered nearly twice (45,000 ÷ 23,000 ≈ 1.96). The example remains illustrative.
FAQ
Does a trading house really need visibility in AI assistants?
Yes, because bankers, counterparties and candidates use them for a first check. Across 20 typical questions, a company with a minimal site is often absent or misdescribed in most answers.
How much does a diagnostic cost in Singapore?
As a 2026 ballpark, about USD 2,300 to 7,700 (SGD 3,000 to 10,000), depending on the number of entities and businesses. Kolonell scopes this diagnostic on quote.
Can we correct false information given by an AI assistant?
You cannot edit an AI assistant directly. You publish clear, consistent information, correct third-party sources where possible, then check progress across several records, often over 3 to 6 months.
Should we publish trading volumes?
Only what the company already publishes officially. No figure should be added to impress AI assistants, and 100% of published facts must remain verifiable.
How important are compliance pages?
They are often the most useful pages: sanctions, know your customer and responsible sourcing, dated and reviewed at least once a year.
Find out how AI assistants describe your firm. We record what AI assistants like ChatGPT say about your trading house today, entity by entity. 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.
