★★★
標竿
實體、答案、結構與證據層皆強；在抽樣查詢中穩定被正確理解，可作品類參考。
AI Search Lab GEO Guide · 2026 夏季刊
GEO Guide list
AI Search Lab rates brands on public five-dimension scores, evidence entry points and fixed buyer questions—how ready they are to be understood, verified and cited in generative search. Every listed brand has its own analysis page, with a breakdown of how AI visibility connects to its business model. Not an ad ranking, and not a shortcut that treats schema as trust.
Why this list exists
When AI starts “recommending,” the market needs a trusted rating language
Users increasingly ask AI which companies are reliable, how to compare them and which fit which businesses. Without public standards, the answer ecosystem collapses into ads, fragmented word-of-mouth and unverifiable claims. The GEO Guide list opens the rating method so brands, buyers and researchers can discuss citability in the same language.
- We rate public digital footprints and AI sample performance—not private deals or paid exposure.
- Stars reflect readiness to be understood and cited correctly—not revenue, popularity or product quality.
- High-risk categories (finance, healthcare) are rated only on information clarity and boundaries—not recommendations or efficacy judgments.
What the stars mean
Three tiers mapped to citability readiness
★★☆
優質
核心頁與實體基礎紮實，多數高意圖問題可被抽取；仍有平台或來源缺口。
★☆☆
入選
已具備可核實基礎與清晰服務定位，值得在品類中被留意；仍需補齊答案或證據層。
Rating criteria
Five signal groups—all publicly reviewable
Each criterion is scored 0–20 for a total of 100; 88–100 is three stars, 72–87 two stars, 55–71 one star. Full method: research methodology. Listing or star upgrades require verifiable strength across most signals—not a single favourable screenshot.
-
01
實體清晰度
中英文名稱、服務範圍、地區、聯絡與作者／組織關係是否一致、可核實。
-
02
可引用答案
高意圖問題是否有直接答案、定義、限制、FAQ，而非只有宣傳口號。
-
03
機器可讀結構
標題層級、內部連結、Schema 與頁面是否易於被檢索及抽取。
-
04
外部證據層
是否存在真實、相關的第三方來源（專業媒體、商家、評論、作者或行業平台）。
-
05
AI 抽樣表現
在固定買家問題集下，品牌是否被提及、描述是否正確；不保證長期引用。
Important limits: The list does not guarantee ongoing mentions or citations on any platform; stars adjust as content, entities, competition and models change. Paying clients are not automatically listed or given higher stars.
2026 夏季刊 · Listed
This season’s GEO Guide list
15 brands, sorted by total score. Open an entry for five-dimension scores, sample questions, evidence, improvement gaps and its GEO commercial conversion path.
SEO／流量獲客
金融服務
公共交通／本地服務
本地零售
高等教育
目的地／公共資訊
金融比較平台
醫療健康
旅遊體驗電商
電子支付／本地生活
餐飲／本地評價平台
航空／旅遊
公用事業
保險
電訊／寬頻
★★★
Higher education
City University of Hong Kong
Hong Kong · 93/100 · Reviewed 2026-08-01
Faculty, programme, research and institutional entity data are rich and verifiable, with strong knowledge-graph signals—an official-source pattern that is unusually easy to cite correctly in education.
- Strong institutional entity
- Deep programme/research pages
- Authoritative third-party corroboration
★★★
Public transport / local services
MTR Corporation
Hong Kong · 92/100 · Reviewed 2026-08-01
Routes, fares, service status and local-intent pages are highly structured, with consistent official entity data—so AI can more readily extract accurate operations information.
- Structured service data
- Local-intent coverage
- Consistent official entity
★★★
Destination / public information
Hong Kong Tourism Board
Hong Kong · 92/100 · Reviewed 2026-08-01
Attractions, itineraries and practical information are presented answer-first, with a clear official identity—so common travel questions are easy to extract as citable passages.
- Answer-first content
- Official identity
- Multilingual coverage
★★★
Electronic payments / local life
Octopus
Hong Kong · 89/100 · Reviewed 2026-08-01
Products, top-up, spend scenarios and card-type explanations are clear, with strong local payment intent coverage. Freshness of policies, fees and campaign terms is key to answer accuracy.
- Very high local intent
- Clear products/scenarios
- Consistent official entity
★★☆
Financial services
HSBC Hong Kong
Hong Kong · 86/100 · Reviewed 2026-08-01
Brand entity is very strong and product/region pages are complete. Even under financial compliance limits it still offers verifiable product explanations—one of the clearer information benchmarks in a high-risk category.
- Strong brand entity
- Complete product explanations
- Multilingual / local pages
The rating reflects only how understandable and verifiable the public digital footprint is; it is not investment or product advice.
View full scores and revenue path →★★☆
Utilities
CLP Power Hong Kong
Hong Kong · 85/100 · Reviewed 2026-08-01
Tariffs, applications, outages and energy-saving information carry official authority. Utility answers must emphasise service territory, safety and the latest tariff tables—and not be confused with generic saving tips.
- Official authority
- Clear service flows
- Local geographic coverage
★★☆
Dining / local review platform
OpenRice
Hong Kong · 84/100 · Reviewed 2026-08-01
Restaurant, district, cuisine and review structure is complete and highly suited to local “where to eat” intent. Ranking transparency, commercial partnerships and data freshness directly affect citation trust.
- Local search intent
- Structured restaurant data
- Dense UGC reviews
The rating is not a recommendation of any restaurant or purchase.
View full scores and revenue path →★★☆
Travel experiences ecommerce
Klook
Hong Kong / multi-market · 83/100 · Reviewed 2026-08-01
Experience pages, locations and booking data are highly structured with broad category coverage. Clearer entity links between brand and individual experience pages would help long-tail questions attribute correctly.
- Structured experience pages
- Local + multi-market coverage
- Complete transactional data
★★☆
Airline / travel
Cathay Pacific
Hong Kong / multi-market · 83/100 · Reviewed 2026-08-01
Route, membership, baggage and service-policy pages are complete with a strong brand entity. Fares, flight status and promotion terms change quickly, so AI sampling must return to live official sources.
- Strong brand entity
- Complete policy pages
- Multilingual service content
★★☆
Financial comparison platform
MoneyHero
Hong Kong · 82/100 · Reviewed 2026-08-01
Dense comparison content and product attributes suit decision queries. Citation trust depends on continuously maintaining update dates, condition limits and source transparency.
- Dense comparison content
- Clear product attributes
- Decision-intent coverage
The rating is not a recommendation of any financial product.
View full scores and revenue path →★★☆
Healthcare
CUHK Medical Centre
Hong Kong · 82/100 · Reviewed 2026-08-01
Specialties, service flows and institutional background are verifiable with a clear information architecture. In high-risk categories, accuracy and professional boundaries come first—no efficacy or recommendation judgements.
- Clear specialty information
- Verifiable institution
- Explicit service boundaries
The healthcare rating assesses only information verifiability and clarity—never clinical outcomes.
View full scores and revenue path →★★☆
Insurance
AIA Hong Kong
Hong Kong · 78/100 · Reviewed 2026-08-01
Product categories, brand and Hong Kong market positioning are clear. Insurance is a high-stakes decision, so answers must keep product facts, eligibility conditions and a “not personal advice” boundary.
- Strong brand entity
- Clear product taxonomy
- High-risk boundary awareness
The rating assesses only information clarity and verifiability; it is not insurance or financial advice.
View full scores and revenue path →★★☆
Telecom / broadband
HKT
Hong Kong · 78/100 · Reviewed 2026-08-01
Broadband, mobile and home-service product lines are clear with high local comparison intent. Plan prices, coverage and contract terms need continuous updates to support reliable citation.
- Local service coverage
- Structured plan pages
- Multi-brand ecosystem
★★☆
SEO / traffic acquisition
TrafficHolic
Hong Kong · 73/100 · Reviewed 2026-08-01
Service scope, delivery logic and Hong Kong market positioning are clear—usable as a verifiable source for traditional SEO and traffic acquisition. GEO-specific answer pages can still go deeper.
- Clear service definitions
- Hong Kong market context
- Verifiable official site
Conflict disclosure: TrafficHolic shares operations and commercial collaboration with AI Search Lab, so this entry should not be read as an independent endorsement. Scores are retained against public fields; readers should prioritise primary evidence and stated limits.
View full scores and revenue path →★☆☆
Local retail
Prince Jewellery & Watch
Hong Kong · 63/100 · Reviewed 2026-08-01
Strong store and category presence with high brand recognition. Strengthening product knowledge, buying criteria and citable FAQs would help move from “recognised” to “explained correctly.”
- Brand recognition
- Store entities
- Category coverage
Nominations and corrections
Want on the list? Prepare verifiable public evidence first
We accept brand self-nominations and third-party nominations. Free nominations enter the editorial candidate pool only. A paid Audit buys research hours, a baseline report and improvement recommendations—not listing, a favourable conclusion or stars.
- Official website and main service／product pages
- Chinese and English brand names, region and contact details
- 3–5 questions you believe buyers will ask AI
- (Optional) Third-party sources you want us to review
FAQ about GEO Directory | AI Search Lab GEO Guide
Is the GEO Directory a paid ranking?
No. Entries are editorial against public criteria. Paying clients are not guaranteed listing or higher stars. Relevant commercial relationships are disclosed on entries.
What do the stars mean?
One star is listed, two is strong, three is exemplary—reflecting readiness across entity clarity, citable answers, machine-readable structure, external evidence, and AI sampling performance.
Does listing guarantee AI citations?
No. Stars describe conditions for being understood and cited. Live answers still shift with platforms, queries, timing, and competition.
How can a brand be nominated?
Submit the official site, service pages, bilingual entity facts, and three to five buyer questions via WhatsApp—or start with a Visibility Audit for a structured baseline.
Are finance or healthcare entries recommendations?
No. Higher-risk categories are scored for information clarity and verifiability only—not investment, medical, or consumer advice.
Do revenue notes prove AI-attributed income?
No. Brand pages infer potential conversion paths from public business models. Only linked referral, qualified conversion, and closed revenue can support attribution claims.
Want a baseline of how AI answers describe your brand?
Request a free Audit