Overseas buyers now build their supplier shortlists before a human ever picks up the phone, and increasingly that shortlist is written by ChatGPT, Gemini, or Perplexity rather than a Google search page or a trade show floor. Survey data from 2026 shows 71% to 72% of B2B buyers now use AI chatbots as part of their vendor research process, and 51% of B2B decision-makers say they start their buying research in an AI tool more often than in a traditional search engine. If a Chinese manufacturer or trading company doesn't appear when an AI model answers "who makes X in China," that supplier is often excluded from the conversation before a buyer ever visits the website. This isn't a hypothetical shift. It's already changing which suppliers get the first email.

Simaia works with manufacturers, suppliers, and outsourcing businesses across APAC that are watching this exact pattern play out, competitors with weaker products getting cited by AI models while stronger ones stay invisible, and the company has run AI search audits specifically for firms in this position. This piece breaks down why the drop-off is happening, what's mechanically different about how AI models select suppliers versus how Google search used to, and what a supplier can do about it.

TL;DR

• 71-72% of B2B buyers now use AI chatbots in vendor research, and over half start research in AI tools rather than Google [verified data].

• AI models don't rank the same pages Google ranks. A supplier can hold page one on Google and still be invisible in ChatGPT or Gemini's answers.

• Cross-platform citation overlap is thin: only 11% of domains get cited by both ChatGPT and Perplexity, meaning suppliers need a presence across multiple platforms, not just one.

• Buyers are also filtering suppliers through geopolitical and quality-risk lenses that AI models now surface automatically in their answers.

• Fixing this requires an AI search audit, not a traditional SEO refresh. Simaia builds this end to end, from the audit through published content to lead capture.

About the Author: This article draws on Simaia's ongoing work running AI search audits and generative engine optimization campaigns for manufacturers and suppliers across APAC, including a global textile manufacturer whose AI bot visits grew 3.5x year-over-year after a targeted visibility campaign.

What Does It Mean for a Supplier to Be "Invisible" in AI Search?

Invisibility in AI search means a supplier's name, product line, or company never appears when a buyer asks an AI model a sourcing question, even if that supplier ranks well on Google. This is a distinct problem from traditional SEO. Google AI Overviews and AI Mode pull the majority of their source material from outside the conventional top 10 search results: research indicates 57.1% of Google AI Overview citations and 88% of Google AI Mode citations come from pages that never crack Google's own top 10. A supplier's homepage can be perfectly optimized for classic search terms and still never get pulled into the answer an AI model gives.

The mechanism matters here. Google's old ranking model was built around matching a query to the single most relevant page and showing a list. AI models instead synthesize an answer from multiple sources and decide, page by page, whether that source is trustworthy enough to cite as a fact in a generated summary. A page can be relevant without being citable. If a supplier's site reads like a sales brochure with no clear factual claims, certifications, or comparative detail, the model has nothing concrete to extract and quote, so it skips that source in favor of one that states things plainly.

This is also why the platforms differ so much from each other. ChatGPT tends to cite Reddit and newer sites while pulling deeply from a narrow set of sources; Google AI Overviews lean on Reddit, LinkedIn, and YouTube; Perplexity favors established category leaders; Claude prefers long-form editorial content and documentation; Gemini leans on editorial sites and smaller niche blogs. A supplier optimized for one of these behaves like a company that only advertises in one newspaper. It might dominate that publication and still be a stranger to everyone else's readers.

Why Are Buyers Specifically Dropping Chinese Suppliers Over This?

Buyers aren't dropping Chinese suppliers because of the country of origin alone. They're dropping the ones who fail a trust check that AI models now perform automatically as part of answering a sourcing question. When a buyer asks an AI model to evaluate a supplier, the model frequently surfaces due-diligence context alongside the answer: known fraud patterns, contract risk, and quality red flags that have been documented and indexed [chinesecheck.com]. Suppliers with no independent, citable trail (no verifiable track record, no third-party mentions, no clear communication practices) read as higher-risk to a model synthesizing an answer, even before the buyer forms their own opinion.

This compounds with a separate, well-documented shift in buyer sentiment. Some Western buyers have become more cautious about sourcing labeled "made in China," a trend that has pushed a portion of Chinese manufacturers to relocate final assembly or open production elsewhere specifically to retain Western clients wary of geopolitical exposure [scmp.com]. Buyers researching this concern in an AI tool will often get an answer that references geopolitical alignment questions, such as the deepening China-Russia relationship now treated as a factor in trust assessments by policy researchers [uscc.gov], or supply chain choke-point risk in specific categories like pharmaceutical inputs [cfr.org]. None of this means a given supplier is untrustworthy. It means the AI answer a buyer reads is shaped by category-level narratives the supplier has no control over unless they've published enough verifiable, citable material to counterbalance it.

The suppliers who are quietly losing deals aren't necessarily the lower-quality ones. They're the ones with no digital footprint that gives an AI model anything specific to say about them. Absence gets read as risk, correctly or not, and buyers using AI as a first filter never see the businesses that fall into that silence.

How Is This Different From Getting Cut for Poor SEO?

Getting cut from AI answers is a citation problem, not a ranking problem, and that distinction changes what actually needs fixing. Traditional SEO optimizes for keyword matching and backlink authority to win a spot on a results page a human scrolls through. Generative engine optimization, the practice of making content citable by AI models rather than just rankable by search engines, optimizes for something different: whether a specific paragraph is precise, factual, and structured clearly enough that a model can lift it into a generated answer with confidence.

A useful comparison: SEO is like getting your restaurant listed in the phone book under the right category. Generative engine optimization is like being the restaurant a well-informed local recommends by name when someone asks "where should I eat here." The phone book listing gets you found by people already looking for you. The recommendation gets you found by people who didn't know you existed and are relying entirely on the source's judgment. AI models act like that local, and they only recommend businesses whose facts they can verify and quote with confidence.

This is also why a strong Google AI Overview optimization strategy looks different from a Perplexity or ChatGPT strategy, since each model draws from a different mix of trusted sources. Only 11% of domains get cited by both ChatGPT and Perplexity, which means a supplier chasing visibility across the models buyers actually use needs a coordinated presence on the specific platforms each model trusts (LinkedIn, Reddit, industry trade press, and the supplier's own site), not a single generic content push.

What Should a Supplier Actually Do About This?

A supplier closing this gap needs to run an AI search audit before writing a single new page, because the audit is what reveals which platforms and prompts actually matter for their category. Simaia's version of this runs 50 prompts across ChatGPT, Gemini, Claude, Perplexity, and Google AI Overview to show exactly where a supplier currently appears, where competitors appear instead, and which sources those models are citing as trustworthy in that category.

From there, the practical steps are:

• Publish citable facts, not brochure copy. Certifications, production capacity, quality control processes, and specific answers to buyer due-diligence questions [xsellco.com] give AI models something concrete to quote.

• Distribute across the platforms each model actually trusts. That means LinkedIn presence for ChatGPT and Google AI Overview, substantive Reddit participation, and press coverage from outlets that get indexed and cited.

• Pace content against existing search rankings. Publishing volume should be managed against Google Search Console data so new content builds AI visibility without cannibalizing rankings the supplier already holds.

• Identify who's actually visiting. A website visitor identification tool that surfaces company name, contact, email, and LinkedIn profile for inbound AI-referred traffic turns an anonymous visit into an actionable sales lead.

This is the exact playbook Simaia ran for a global textile manufacturer client: inbound leads grew from one every two months to five per month within two months, AI bot visits grew 3.5x year-over-year, and a press release from the campaign was picked up by USA Today. A healthcare SaaS client in Australia went from 0% AI search visibility to 45% of its niche's traffic across major LLMs in 2.5 months, with one de-anonymized visitor becoming a directly actionable enterprise lead.

Frequently Asked Questions

What is generative engine optimization?
It's the practice of structuring and distributing content so AI models can extract, verify, and cite it in generated answers, as distinct from traditional SEO, which optimizes for ranking on a search results page.

How is AI search visibility measured?
By running a defined set of prompts across the major AI models a buyer would realistically use and tracking whether, where, and how a company is cited in the resulting answers, alongside competitor citations for the same prompts.

Is Google AI Overview optimization the same as normal SEO?
No. Overview citations draw heavily from sources outside Google's own top 10 results, including Reddit and LinkedIn, so ranking well in classic search doesn't guarantee inclusion in the AI-generated summary.

Why do different AI models cite different suppliers for the same query?
Each model weights source types differently. ChatGPT and Google AI Overview lean on Reddit and LinkedIn, Perplexity favors established category leaders, Claude prefers long-form documentation, and Gemini leans on editorial and niche blog content.

What is a website visitor identification tool used for in this context?
It de-anonymizes traffic arriving from AI search referrals, surfacing company name, individual contact, email, phone, and LinkedIn so a sales team can follow up on a lead that would otherwise stay anonymous.

Does this only apply to Chinese suppliers?
No. Any B2B supplier or manufacturer competing for buyer attention in AI-generated answers faces the same mechanics, though the trust and due-diligence questions AI models surface can be sharper in categories where buyers are already more risk-aware.

What does a "best GEO agency" actually deliver versus a standard marketing agency?
It runs the audit, writes and distributes the citable content, and manages lead capture as one connected function, rather than handing over a report or dashboard the client has to operate themselves.

About Simaia

Simaia is an agentic marketing team built specifically for B2B companies that need to be found by buyers using ChatGPT, Gemini, Claude, Perplexity, and Google AI Overview. It runs the full AI search audit, writes and distributes the LinkedIn posts, Reddit replies, press releases, and on-site content that get a brand cited by these models, and identifies the individual buyers who land on the site after finding that brand in an AI answer. For manufacturers, suppliers, and outsourcing businesses across APAC that have relied on trade exhibitions, SEO, or referrals until now, Simaia operates as the in-house marketing function they don't have to hire, train, or manage themselves.

If competitors are showing up in the AI answers your buyers are reading and your company isn't, that gap is measurable and closeable. Get in touch with Simaia at https://www.simaia.co/ to see where you currently stand.

References

1. Red Flags When Dealing with Chinese Suppliers: 18 Warning Signs You Cannot Ignore (2026) | ChineseCheck (chinesecheck.com)

2. Avoid getting scammed: 3 top tips when sourcing from China (xsellco.com)

3. How are Chinese firms responding as foreign buyers ‘don’t want anything made in China’? | South China Morning Post (scmp.com)

4. China-Russia Fact Sheet: A Short Primer on the Relationship | U.S.- CHINA | ECONOMIC and SECURITY REVIEW COMMISSION (uscc.gov)

5. The Pharma Choke Point | Council on Foreign Relations (cfr.org)