The verdict in three sentences
For a commodity trading house in Tokyo, the website exists to reassure trade finance banks, counterparties, regulators and recruits, not to sell a product. SEO still matters for the brand, offices and recruiting, while AEO/GEO answers a newer question: what do AI assistants like ChatGPT say about your governance, compliance and responsible sourcing when a credit analyst or a journalist asks them? The right split depends on how well known you already are and on the quality of the compliance pages you have actually published, and it is measured through qualified leads on the SEO side and a dated record of answers on the AI side.
What each lever really targets for a trader
A trading house has no shopping cart. Its high-value visitors are a trade finance officer at a Japanese or Singaporean bank, a counterparty's KYC team, an NGO checking a supply chain, a bilingual Japanese and English graduate. They type few generic queries and increasingly ask full questions to assistants.
| Criterion | SEO (Google) | AEO/GEO (AI assistants like ChatGPT) |
|---|---|---|
| What it targets | Brand queries, offices, careers, "commodity trader Tokyo" | Due diligence questions: "does this trader publish a sanctions policy?" |
| Time to effect | 4 to 9 months on low-competition queries | 2 to 6 months to see answers shift, with no certainty |
| Tokyo market budget (2026 ballpark) | JPY 300,000 to 900,000 per month, about USD 2,000 to 6,000 | One-off diagnostic JPY 600,000 to 1,500,000, about USD 4,000 to 10,000 |
| Metrics | Search Console impressions and clicks, applications, inbound contacts | Share of questions where the firm is cited, accuracy of the facts repeated |
| Main risk | Worthless traffic if you chase product keywords | Wrong or outdated compliance answers that are slow to correct |
| Key content | Office, business line and careers pages in Japanese and English | Governance, KYC, sanctions, ESG report as actually published |
These budgets are 2026 ballpark figures drawn from typical quotes by Tokyo agencies, converted approximately at around JPY 150 to USD 1.
Where to invest, situation by situation
There is no single answer. A young trader that has just opened an office in Marunouchi does not face the same problem as an established house already named in the trade press.
| Situation | Suggested priority | Why |
|---|---|---|
| Young firm, rarely cited | Brand SEO and institutional pages first | AI assistants draw on existing sources: without solid pages there is nothing to pick up |
| Established brand, often cited | AEO/GEO first | The risk is inaccuracy: closed offices, former executives, stale figures |
| Regulated content (sanctions, KYC, minerals) | Compliance review before any lever | A loose sentence repeated by an AI keeps circulating |
| Limited budget (under USD 3,000 a month) | One-off AI diagnostic, then careers-focused SEO | A baseline record avoids investing blind |
| Large graduate intake | Japanese SEO on careers pages | Candidates still rely heavily on Google and job boards |
| Raising credit lines | AEO/GEO on governance and published statements | Credit analysts increasingly query assistants |
On the Japanese side, data protection falls under the APPI (Act on the Protection of Personal Information), overseen by the Personal Information Protection Commission. It mainly affects your application and contact forms. Sanctions and export controls fall notably under the Foreign Exchange and Foreign Trade Act. This is not legal advice: your lawyers sign off each page before it goes live.
How to measure without fooling yourself
SEO is measured with familiar tools: Search Console for impressions and queries, analytics for journeys, the CRM for contacts and applications. For a trader, fifteen qualified contacts from banks or counterparties are worth more than 10,000 visits.
AEO/GEO is measured with a dated record: the same list of questions, put to AI assistants like ChatGPT at fixed intervals, with each answer archived. You track three things: is the firm cited, are the facts right, which sources are picked up. No serious provider promises that an AI will recommend you, and none promises a Google ranking.
A serious diagnostic usually covers:
- the list of questions your audiences really ask (banks, counterparties, NGOs, candidates);
- the state of the pages that act as sources: governance, compliance, ESG, offices, careers;
- consistency between the Japanese and English versions;
- factual gaps found in current answers;
- a costed recommendation on the split between SEO and AEO/GEO.
Kolonell's work on this topic is priced on quote, after that diagnostic.
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: Kanda Resources Trading, a mid-sized metals trader based in Chiyoda, has an annual budget of JPY 9,000,000, about USD 60,000. The head of communications is torn between putting everything into SEO or everything into AI visibility.
A baseline record of 30 questions shows the firm cited in only 4 answers, about 13%, and two answers mention a closed office. He splits the budget: 60% to SEO (JPY 5,400,000, about USD 36,000) on careers and office pages, 40% to AEO/GEO (JPY 3,600,000, about USD 24,000) on governance and source corrections. The record is repeated monthly: 30 questions over 12 months gives 360 archived answers, a solid basis for next year's decision.
FAQ
Is SEO still enough for a trading house in Tokyo?
For careers and brand, often yes, with visible effect in 4 to 9 months. For bank and NGO due diligence, questions increasingly go through AI assistants, hence the value of a dated record.
What does an AEO/GEO diagnostic cost in this market?
Expect a 2026 ballpark of JPY 600,000 to 1,500,000, about USD 4,000 to 10,000, depending on the number of questions and languages. At Kolonell, this work is on quote.
Do we need to publish volumes or counterparties to be cited?
No. Publish only what your board and lawyers have approved: governance, policies, the existing ESG report. AI assistants repeat approximations too, so less but accurate is better.
How many questions should we track each month?
Between 20 and 40 is enough for a trader, spread across banks, counterparties, NGOs and candidates. Beyond that, the analysis costs more than it returns.
How soon is there a first useful finding?
The baseline record is ready in 2 to 3 weeks. Shifts in answers show over 3 to 6 months, with no guaranteed outcome.
Find out how AI assistants describe your firm. A dated record of 30 questions in Japanese and English shows what your banks and counterparties already read about you. 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.
