2026 Greater China enterprise GEO and AI search landscape survey: first-hand data on AI citation rates

In Q1 2026, we took a set of industry questions and ran them through ChatGPT, Perplexity, and Gemini, one by one, to track which brands appeared in each response.

When asked "How do I compare loan options for small and medium businesses in Taiwan?" an AI returns two brand names, two media articles, and a 2021 forum post. One client appears nowhere. They spend seven figures annually on SEO and rank on Google's first page, but the AI doesn't mention them.

This pattern held across our entire survey.

Greater China GEO survey: our methodology

The "Greater China GEO Survey" sampled 214 brands across Taiwan, Hong Kong, and mainland China in six industries: financial services, healthcare, manufacturing, retail, logistics, and automotive. For each brand, we developed 8 to 12 questions that real customers would ask. We ran each question in Traditional and Simplified Chinese across three major AI assistants, yielding roughly 5,900 total queries.

Our citation criteria were strict: a brand only counts as cited if its name or website appears as an explicit mention or link. Generic references within industry discussion don't count.

Several limitations matter. AI responses are non-deterministic; asking the same question three times can yield different answers, so we used majority voting. Our queries focused on B2B and high-ticket consumer decisions, so results don't represent all categories. This is a snapshot, not a census.

First-hand data: only 12% of brands ever cited

Only 12% of surveyed brands have ever been cited even once in AI answers to their core category questions. That means more than 80% are invisible when their customers use AI to decide. It's not a ranking issue. These brands don't appear in the response at all.

The gaps vary by industry:

Retail and financial services lead not because they understand AI better, but because they've accumulated third-party content around them: comparison sites, media reviews, social discussion. Manufacturing and logistics lag because there's less public Chinese-language content about them online.

Three patterns deserve attention. Citations come from third-party sources 71% of the time: media, Wikipedia, forums, comparison sites. AI prefers what others say about you, regardless of your website's polish. Most marketing leaders (63%) haven't checked how their brand appears in AI responses, not from apathy but from lacking baseline measurement and tools. Traditional Chinese reveals a content gap: for the same brands, Simplified Chinese queries generate 1.8x more citations. Chinese-language training data is sparse; Traditional Chinese is even sparser. For Taiwan and Hong Kong brands, this signals both a warning and an opportunity before positions solidify.

GEO ROI isn't about traffic

Clients often ask how to calculate GEO ROI, expecting a straightforward ROAS figure like advertising provides. The data isn't mature enough for that yet. Track three earlier-stage metrics that show more direct attribution: citation rate (percentage of category questions that mention you), citation share (your citations relative to competitors in any given query), and source composition (whether citations point to your site or third-party sources).

These metrics move before actual AI-driven referral traffic appears. From a brand's first serious GEO work to stable citation, the median duration we've seen is about 10 weeks. That's faster than traditional SEO but not fast enough for monthly results. Anyone promising ChatGPT recommendations in two weeks isn't being realistic.

We'll rerun this survey every six months to build a long-term curve for Traditional Chinese. Few organizations in the Traditional Chinese-speaking world generate baseline data like this. Without a baseline, measuring progress is impossible. Tenten's approach starts with measuring your current citation rate, then works backward to identify content gaps and where AI should encounter you. Demo results look compelling, but what matters is whether customers see your brand cited when they actually need it.