GEO Blog
收录 GEO 相关文章、案例与观点，内容持续更新，方便延伸阅读。
2026/08/27 01:03
Chinese AI models hedge on supplement queries, concentrating citations on blue-hat-registered brands like By-Health and Swisse — leaving social-commerce insurgents invisible.
2026/08/26 01:03
Xiaohongshu — the most-cited consumer source for Doubao, DeepSeek and Qwen — launched its own RedOne model and standalone AI app. The citation source is becoming an answer engine, splitting the ROI of brand seeding.
2026/08/25 09:00
KFC linked 13,000+ stores to Qwen's brand-agent platform, splitting China GEO into owned-agent vs. citation tracks.
2026/08/24 01:06
VTech's China sales are shrinking while iFlytek's Alpha Egg and other AI-native toys dominate AI search answers.
2026/08/23 01:05
Alipay's AI shopping agent completed 120M+ payments in one week. What autonomous AI checkout means for brand GEO in China.
2026/08/22 09:00
DeepSeek skews 42% Gen Z; Doubao has the highest senior user share of any major LLM. A category-by-category map of which Chinese AI model to prioritize.
2026/08/21 01:10
L'Oréal leads AI citation share at 6.72%; Estée Lauder ranks #18. In China, the same ingredient-transparency bias is reshaping who gets recommended.
2026/08/20 01:04
Doubao hit 382M MAU in June 2026 but earns under $1M/day from consumers. The real monetization path is e-commerce and B2B, not subscriptions.
2026/08/19 01:03
Doubao quoted a MacBook Air M3 at RMB 2,550 in a five-model shopping test. With Qwen-Taobao and Doubao's commerce loops now closed, brand fact accuracy matters as much as visibility.
2026/08/18 01:03
Doubao shipped free full-duplex audio-video AI to ~345M users on Aug 6, 2026. Spoken answers name one brand, not five — and text-based visibility scores no longer measure what brands actually get.
2026/08/17 01:03
Kuaishou's OneRec now serves 25% of feed traffic at 10.6% of cascade cost, lifting local-services GMV 21%. China's behavior layer is generative too — and no citation-based GEO tactic touches it.
2026/08/16 01:03
Doubao and Kimi now sell paid tiers where deep research reads dozens to hundreds of sources instead of a handful. Free-tier scores reward brand fame; paid-tier answers reward retrievable evidence — and almost nobody is measuring both.
2026/08/15 01:03
Alipay's AI Pay processed 120 million agent-executed transactions in one week — with almost no brand citations shown. Qwen's commerce surface reaches 300M users versus 166M in chat. Brand visibility scoring now covers only half the picture.
2026/08/14 01:03
New vendor data suggests every 10pp gain in authority citation rate lifts AI recommendation rates 22–28% across Doubao, Qwen and DeepSeek. What actually moves the needle — and why to act before Doubao's Q4 ad auction opens.
2026/08/13 09:00
Tencent's WeChat AI agent skips screen-tapping, calling mini-programs as structured interfaces -- making mini-program readiness a new brand GEO factor.
2026/08/12 01:03
Kimi's consumer MAU fell for 4 straight quarters, yet Moonshot just raised $3.5B at a $35B valuation and open-sourced its 2.8T model. Here's what the split means for GEO.
2026/08/11 01:03
ByteDance will open a standardized Doubao ad system in Q4 2026, ending the organic-only era in Chinese AI search. Brands scoring 30-60 across the six models are the most exposed.
2026/08/10 01:03
Baidu, Tencent and ByteDance each own both an AI assistant and their own encyclopedia. That vertical integration, not prompt phrasing, explains most cross-model brand score gaps in China.
2026/08/09 01:04
Wire press releases drive 0.04% of AI citations — and Doubao and Qwen never index them at all.
2026/08/08 00:00
StepFun's models already run on 60% of top China phone brands (42M+ devices); its new STEPX agent phone plus IDC's 147M-unit 2026 forecast signal a shift from AI citation to AI-executed transactions.
2026/08/07 01:05
Doubao's Seedream/Seedance models generate AI-native product visuals for its 227M+ MAU shopping loop — brands not using them may be invisible to its own recommendation logic.
2026/08/06 01:02
New data: DeepSeek and Qwen pick the same luxury brands 70% of the time, but cite matching sources only 27% of the time.
2026/08/05 00:00
Doubao leads retail banking AI queries at 345M MAU while DeepSeek owns compliance workflows in China.
2026/08/04 01:05
Qwen3.8-Max ships open weights next week with a 1M-token context — multiplying which AI surfaces run Qwen and rewarding deep brand content.
2026/08/03 01:04
Hermès leads Similarweb's global luxury AI visibility rankings, but that score doesn't predict performance on Doubao, DeepSeek, Kimi, or Qwen — China's AI systems cite an almost entirely different evidence layer.
2026/08/02 01:05
RED sees 70% of its ~350M users search in-app, yet Doubao and DeepSeek rarely cite it. Why that gap now splits your China budget in two.
2026/08/01 01:04
Qwen's MAU grew 5,792.9% YoY in June 2026 while per-user session duration fell 14.1%. QuestMobile's H1 data shows why Doubao's 382M-user lead over Qwen is wider than the ranking implies — and why MAU is the wrong denominator for China GEO budgets.
2026/07/31 01:03
China now has a 17-item GEO service standard (AIIA/T 0277-2026) in a market that grew ~11x to RMB 3B. It is a voluntary association standard being marketed as law — here is how brand teams should use it.
2026/07/30 01:00
Zhipu pledged zero monetization for two years while Doubao already runs an AI commerce loop. Here's what China's 2026 AI funding rush means for GEO budgets.
2026/07/29 09:00
ByteDance's Doubao and Honor just shipped OS-level AI agents on phone hardware, the same week China banned DIY chatbots. Here's what system-level agents mean for brand visibility.
2026/07/28 01:04
Kimi K3's multimodal input and 1M-token context change what brand content Kimi trusts and cites in China's AI search.
2026/07/27 01:03
China's GEO agency market claims ¥18.6B+ scale and 70% enterprise penetration — but vendor self-promotion is polluting the very AI engines (Doubao, DeepSeek, Qwen) it optimizes. A buyer's guide for international brands.
2026/07/26 01:03
China's July 15 rules forced Doubao and Qwen to disable humanlike agent features — but exempted the Q&A surfaces where brand citations happen. GEO signals now decide an even larger share of brand discovery.
2026/07/25 01:04
DeepSeek shipped its most capable model, V4 GA, in July 2026 — the same month its consumer MAU fell 20.3% to 130M while Doubao hit 382M. Why capability and reach have split, and how brands should reallocate China GEO budget.
2026/07/24 01:03
DeepSeek's app fell to 129.8M MAU (-20.3% YoY) in June 2026 while Doubao hit 382.3M. Here's how consumer brands should reallocate China GEO budget.
2026/07/23 01:03
WeChat's new native AI assistant Xiaowei (1.4B user base) launched as Doubao lost 11.7M MAU post-paywall — shifting China brand discovery from AI apps to super-app layers.
2026/07/22 01:03
China's AI-native apps hit 499M MAU in May 2026, up 85.4% YoY. Doubao's 382M lead reshapes how brands should weight GEO budgets across models.
2026/07/21 01:03
Chinese users now triage queries across DeepSeek, Qwen and Doubao by intent. That routing habit means one blended AI score hides which buyer you're losing.
2026/07/20 01:04
Arc'teryx tops hubGEO's outdoor apparel scores across China's six AI models while The North Face and Patagonia trail. The reason is source clarity, not brand awareness.
2026/07/19 01:03
Qwen weighs Tmall ratings, Doubao weighs Douyin buzz, Kimi weighs research documents. In 2026 the same brand can score 70 in one engine and 20 in another — because each model trusts a different kind of proof.
2026/07/18 01:03
QuestMobile's Q1 2026 data shows Doubao (344.5M MAU) and peers growing via silver-haired and lower-tier users — a demographic shift that quietly reshapes which brands Chinese AI models recommend.
2026/07/17 01:04
DeepSeek-V4 launched mid-July 2026 with a 1M-token context, 73% lower inference cost, and peak-valley API pricing. Here's how each change rewires brand citation dynamics in Chinese AI search.
2026/07/16 01:03
Royal Canin and Purina still top Chinese AI recommendations, but Myfoodie, Wanpy, and fresh-food challengers are closing the citation gap across Doubao, Kimi, DeepSeek, and Qwen.
2026/07/15 01:04
Doubao users query 2.8x more than Qwen's. Weighting China AI visibility by session frequency, not MAU, collapses Qwen to 12% and lifts Doubao to 69% of brand-citation exposure.
2026/07/14 09:00
China's $110B infant formula market is one category where Doubao, DeepSeek, and Qwen hedge instead of recommending brands directly — here's why, and what foreign maternity brands should do about it.
2026/07/13 09:00
Tesla now runs Doubao and DeepSeek side-by-side in its China-market Model Y. For foreign automakers, the in-car assistant is a new, untested GEO surface.
2026/07/12 09:00
Testing DeepSeek, Kimi, Doubao, and Qwen on premium appliances: all four recommend Dyson and IQAir over Haier or Midea.
2026/07/10 01:06
Fast food averages 79 across China's six AI models, luxury 75.7, beauty just 55.8. hubGEO's category benchmark reveals which sectors have the highest AI-visibility floor.
2026/07/09 01:00
ERNIE 5.1 hit #1 domestically (#4 globally) on LMArena Search Arena and crossed 200M MAU. Its Baidu Baike/Zhidao-rooted citation logic rewards brands differently than Doubao or Kimi.
2026/07/08 01:06
Sony averages 75 across China's six AI models; Bose averages 36. Real brand-score data shows why single-category audio brands lose GEO ground to multi-vertical rivals.
2026/07/07 01:03
A 2026 Yiou Think Tank report finds alcohol is among the lowest-maturity GEO categories in China — Moutai and Hennessy dominate brand value rankings but are barely cited by Doubao and DeepSeek.
2026/07/06 00:00
Doubao (159M MAU) and DeepSeek (146M MAU) now hold 88.4% of China's AI assistant market — but they cite brands in opposite ways, so budget can't be split evenly.
2026/07/05 01:04
Qwen DAU jumped 727% in a day during China's RMB 45B AI subsidy war, then halved. Here's what the surge actually meant for brand GEO strategy.
2026/07/04 09:00
SMZDM appears in 80-99% of DeepSeek and Qwen luxury recommendations, but that dominance collapses at the Birkin-tier — where Vogue Business and auction archives take over instead.
2026/07/03 01:04
Doubao's 300M+ MAU travel-planning surge favors OTA aggregators over global hotel and airline brands. Here's what the AI travel workflow gets wrong for hospitality marketers.
2026/07/02 08:00
KFC earns a perfect 100 in China's AI brand visibility while newer global chains score near zero. hubGEO data reveals how delivery platform integration, decades of localized content, and Douyin presence drive AI citations for F&B brands across Doubao, DeepSeek, and Qwen.
2026/07/01 01:05
Doubao dominates headlines at 260M MAU, but Baidu ERNIE leads enterprise API. For brand marketers, China's AI has bifurcated into consumer and enterprise layers — and each demands a completely different GEO playbook in H2 2026.
2026/06/30 01:03
Kimi's MAU collapsed 77% from 36 million to 8.3 million in 18 months, pushing it from #2 to #8 in China's AI market. Here's how to recalibrate your China GEO strategy for 2026.
2026/06/29 01:04
Alibaba's Quark has surpassed Doubao as China's most-used AI app — and its deep integration with Tmall's commerce data creates a citation advantage that fundamentally changes how brand GEO scores are calculated in China's largest retail ecosystem.
2026/06/28 01:05
Doubao's 345M MAU base now routes beauty recommendations directly to Douyin purchase links. Here's what this AI-to-commerce loop means for international skincare brands in China.
2026/06/27 05:29
The same brand scores 68 on Kimi and 11 on Wenxin. Understanding why China's six AI platforms cite brands differently is now a core GEO intelligence challenge for international marketers.
2026/06/27 00:03
DeepSeek is now integrated into WeChat Search, giving it access to 800M users and WeChat's exclusive content ecosystem. Here's what brand marketers must know about citation logic and GEO strategy.
2026/06/25 00:00
In consumer electronics — China's highest-consideration purchase category — AI recommendation rates diverge sharply from market share. hubGEO's June 2026 data shows why Huawei dominates Doubao while Apple underperforms, and what the 60% pre-purchase AI research figure means for foreign brands.
2026/06/24 01:03
Doubao has 345M MAU. Kimi has 8.3M. That 41× gap is reshaping where China's AI brand discovery actually happens — and most international brand GEO budgets haven't caught up yet.
2026/06/22 08:00
Doubao's 260M+ monthly users see responses shaped by ByteDance's training data — brands with Douyin presence achieve nearly 100% citation coverage in China's #1 AI platform, while those without it are structurally invisible.
2026/06/21 06:00
Doubao leads China AI at 345M MAU — 41x Kimi's 8.3M. New Q1 rankings reshape how brands should allocate China GEO budgets.
2026/06/20 01:04
Doubao hit 345M MAU, Qwen surged to 166M after Alibaba's brand unification, and Kimi collapsed from 36M to under 15M. Here's how to reprioritize your China AI brand visibility strategy.
2026/06/19 01:03
Doubao, DeepSeek, and Kimi now reach 462M monthly users combined—and Xiaohongshu content is absent from their brand citations. Here's what that means for your China visibility strategy.
2026/06/18 01:03
Doubao and Qwen deployed full-scale AI commerce integrations during China's 618 festival, creating two distinct brand discovery architectures. Here's what it means for your GEO strategy.
2026/06/17 05:44
Doubao lost 6M users when subscription pricing leaked—then pressed forward anyway. Here's what ByteDance's four-tier model and Douyin commerce integration mean for how brands get recommended across China's 260M-user AI platform.
2026/06/16 01:03
Doubao hits 260M MAU and launches AI e-commerce, Quark surges to 180M — the China AI search map has shifted, and brand GEO strategy must follow.
2026/06/15 01:04
Doubao's 'Help You Choose' feature launched in May 2026, turning China's largest AI app (345M users) into a brand discovery-to-purchase engine. Here's what it means for your China AI visibility strategy.
2026/06/14 01:04
Nike scores 82/100 in Chinese AI visibility while Lululemon — despite 47% China revenue growth — scores just 44. We break down why legacy content depth beats commercial momentum in China's AI search era.
2026/06/13 01:04
31% of Chinese enterprise decision-makers now use AI to research vendors before contacting sales. Here's how DeepSeek, Qwen, and Doubao decide which B2B brands make the shortlist.
2026/06/12 08:00
Our tracking of 7 major auto brands across 6 Chinese AI models reveals a clear pattern: German legacy brands dominate, while newer entrants and niche luxury are invisible. Here's what the data means for your GEO strategy.
2026/06/12 08:00
Among the 8 luxury brands we track in Chinese AI, the gap between leaders and laggards is absolute. Understanding why reveals the structural drivers of luxury GEO — and what brands on the wrong side must do.
2026/06/12 08:00
Nearly a third of the brands we track score zero across all Chinese AI models. This isn't bad luck — it's a predictable result of specific, fixable content gaps. Here's our diagnosis and the 90-day action plan for zero-score brands.
2026/06/12 08:00
The beauty category shows the most polarized China AI visibility scores of any sector we track. Hero products, e-commerce integration, and decades of brand investment separate winners from invisible laggards.
2026/06/12 08:00
Sixteen brands in our database achieve a perfect score of 100 across all six Chinese AI models. We analyzed what they share — and built a blueprint any brand can use to move toward perfect China AI visibility.
2026/06/12 00:00
ByteDance's Doubao crossed 260 million monthly active users in Q1 2026, surpassing Baidu ERNIE Bot for the first time and rewriting the rules of brand visibility in Chinese AI search. Here is what the shift means for foreign and domestic brands building a GEO playbook for the next twelve months.
2026/06/11 00:00
China now has six competing AI answer engines — DeepSeek, Kimi, Doubao, Qwen, Yuanbao, and Baidu ERNIE — each with different sources and citation logic. Here's why that fragmentation changes everything about GEO strategy for brands in China.
2026/06/10 08:00
DeepSeek hit 541 million monthly visits in May 2026 while Baidu stages a multi-product AI comeback. Here is what the shifting power dynamics in China's AI search ecosystem mean for brand visibility and GEO strategy.
2026/06/06 05:31
在哪些平台發文，才能被中文AI引用？本文整理豆包、DeepSeek、通義千問、腾訊元寶、百度文心一言五大模型的GEO引用來源數據，幫助品牌精準選擇發文渠道。
2026/06/06 00:00
Doubao, Kimi, Qwen, DeepSeek, and Baidu ERNIE each use different recommendation logic. Here's why a one-size-fits-all GEO strategy fails in China's 2026 AI search landscape - and what to do instead.
2026/06/04 00:00
Doubao surpassed Baidu in MAU for the first time in Q1 2026, Kimi K2.6 topped the coding benchmarks, and DeepSeek V3.2 deepened its grip on enterprise. For GEO practitioners, this three-way competition is fundamentally reshaping which signals matter for brand recommendations in China's AI search landscape.
2026/06/03 00:00
With 150 million MAU and WeChat-native distribution, Tencent Yuanbao is a distinct GEO surface that rewards a different optimization playbook than DeepSeek, Doubao, or Kimi. Here's what brands need to know in 2026.
2026/06/02 08:00
China's GenAI user base hit 602 million in 2025 — and over 70% of searches now end without a single click to any website. Here's what that means for your GEO strategy on Doubao, DeepSeek, Kimi, and Qwen.
2026/05/31 08:00
China's largest local-life platform just launched a GEO marketing portal. From a ¥10M procurement to building its own AI visibility infrastructure — here's what Meituan's move signals for brands operating in the Chinese market.
2026/05/31 00:00
Kimi K2.6 and DeepSeek V4 have cracked the global AI rankings at #1 and #2. For brand marketers, this power shift rewrites the rules of AI visibility in the world's largest digital market.
2026/05/29 08:00
The same brand query, eight different AI platforms, dramatically different results. After testing hundreds of brands across China's major AI models, here's what drives the recommendation gap — and what you can do about it.
2026/05/28 08:00
Cross-industry audit data from hundreds of Chinese brands reveals a stark picture: most brands are missing from AI-generated recommendations entirely, hallucination rates reach 58% for smaller brands, and the window to establish first-mover advantage is closing.
2026/03/05 08:00
China's AI search ecosystem is unlike anything in the West. Six major AI platforms, distinct citation patterns, and no Western GEO tools cover it. Here's what you need to know.
2026/01/20 08:00
很多品牌發現自己在 Google 排得很好，但在 ChatGPT 的回答裡完全不存在。這不是偶然，而是有跡可循的結構性問題。
2026/01/15 08:00
中文 AI 搜索不只有一個平台，各自的引用邏輯、用戶群體與重點場景都不同。這篇幫你理解哪個平台值得優先布局。