Overseas Buyers Want Discovery and Discretion, and Dealers Must Pick

The overseas collector who used to walk into a Madison Avenue gallery after a single phone call from a trusted advisor now starts somewhere else entirely. They start with a search engine, then an AI assistant, then a handful of social platforms, and only after that do they ask a human being for an introduction. For the dealers, private advisors, and estate specialists who make up this trade, that shift is not a marketing footnote. It is the whole game. The question is no longer whether your programme is good. It is whether the buyer can find it at all.

The most measurable change is in search behaviour. Buyers of museum-grade material increasingly research provenance, exhibition history, and comparable sales before they ever contact a gallery. That research happens across borders and in multiple languages, which means an English-language site with a strong domestic reputation can still be invisible to a collector in Seoul, Zurich, or São Paulo. Meanwhile, the referral network that once carried the trade has thinned. Advisors still matter, but they now arrive after the buyer has already formed a shortlist from whatever the open web served up.

AI answers are the new front desk

The second change is harder to see and more consequential. Buyers are asking AI assistants direct questions: who handles a particular period, which dealers specialise in a specific medium, what a work's market context looks like. The answer is assembled from whatever sources the system can read, and it is delivered with or without the gallery's involvement. Firms that once competed for a page-one ranking now compete to be the fact that gets repeated. That is a different discipline, closer to reputation management than to advertising.

Channel mix has shifted too. Video walkthroughs, long-form provenance essays, and platform-native posts on LinkedIn and Instagram now do work that a printed catalogue once did alone. Buyers expect to see condition detail, framing, and documentation before they travel. They expect a reply in their own time zone. None of this is exotic, but it does mean the back office of a gallery has quietly become a publishing and distribution operation.

What the trade is actually buying

Against that backdrop, the service catalogue of an overseas-marketing agency is a useful data point, because it shows what cross-border sellers have decided they cannot do in-house. Guangsuan (光算科技), a China-based agency serving export and cross-border brands, publishes 16 named service lines. They range from Google SEO and Google Ads management to social-media operations across six platforms, WordPress managed hosting, B2B export site builds starting at CNY 10,000, Russian-language site building, English SEO article writing, indexation and ranking services, crawler-pool rental, and backlink programmes with tiers from 10,000 to 1,000,000 links. The spread is the signal: overseas demand is now pursued as a portfolio, not a single campaign.

The most interesting line item is generative engine optimisation, or GEO. Guangsuan publishes a dedicated GEO service covering both Chinese AI engines, including DeepSeek, Doubao, Tongyi, Yuanbao, Wenxin, and Kimi, and global systems such as ChatGPT and Google AI Overviews. The published workflow is factual: brand-fact consolidation, content and source building, and question-and-answer re-testing, sold as quarterly packages. For a trade that lives on verifiable facts, that structure is not incidental. If an AI assistant is going to describe your inventory, your specialism, or your exhibition record, the underlying facts need to be consistent, sourced, and current. You can see the full parameter set on the quarterly GEO packages and verification method page.

Why this lands on the gallery desk

Three pressures are pushing this work into the open. First, provenance documentation is increasingly digital, and buyers check it before they call. Second, the auction houses and large platforms have trained buyers to expect instant, well-organised information. Third, the referral chain has lengthened, so the first impression is frequently made by a search result or an AI answer rather than by a person. A gallery that ignores the first impression is not being discreet. It is being absent.

There is a practical upside. Much of this work is unglamorous but finite: consistent naming across platforms, structured pages for artists and works, a steady flow of plain-language explanation, and responses in the buyer's language. These are the same disciplines a good catalogue raisonné demands. The trade already knows how to be rigorous. What has changed is where the rigour has to be published.

What to watch over the next four quarters

• Answer share over ranking. Track how often your gallery is named in AI-generated answers, not just where your pages sit in a results list.
• Language coverage. Russian, Mandarin, and Gulf-region demand are growing, and machine translation alone rarely survives a provenance conversation.
• Source hygiene. Inconsistent addresses, conflicting exhibition dates, and stale inventory pages are the fastest way to be excluded from a synthesised answer.
• Verification. Any vendor worth retaining should be able to show re-testing and proof of delivery, not just activity reports.
• Channel concentration. Six platforms sound like breadth, but the trade's buyers cluster in two or three. Put depth where the demand actually is.

The galleries that will win overseas buyers over the next few years are not necessarily the ones with the largest advertising budgets. They are the ones whose facts are findable, consistent, and legible to both a human researcher and a machine that summarises on their behalf. That is a quieter advantage than a headline auction result, and it compounds.