1 / 30
Winning in China's
AI Search Revolution
From Invisible to Recommended: Pilot Results & 2026 Strategy
March 9, 2026  |  Confidential
00
Agenda
01
The Shift China's AI search revolution is here — and it changes everything
02
The Gap MitoQ is nearly invisible in AI recommendations — competitors are not
03
The Proof Our 2-week pilot cracked the code — methodology validated
04
The Plan Full-year strategy to make MitoQ the brand AI recommends

MitoQ GEO Intelligent Strategy Architecture

North Star Strategic Goals
🎯 Strategic Goals
☆When consumers ask AI "which mitochondrial supplement is best," MitoQ is the #1 recommendation
☆5/7 KPI keywords ranked in Doubao + Yuanbao Top 6 — Q2 target already achieved in pilot
☆Build GEO as long-term brand recognition infrastructure, not a one-off marketing campaign
☆Establish AI citation moat during the window before competitors enter — capture first-mover advantage
· · · · · · · · · · · · · · · · · ·
Value Proposition Value Layer
⭐ Value Propositions
  • AI Cognitive Positioning: As AI becomes the gateway for purchase decisions, brands must appear in AI answers
  • High-Intent Traffic: AI-referred traffic converts at 7% vs 5% traditional search — late-funnel quality traffic
  • Scalable Content Production: AI-assisted 10,000+ semantically optimized articles — cost does not scale linearly
⚠ Pain Points
  • Zhihu/Xiaohongshu/Douyin = 0 AI citations; Sohu/SMZDM/NetEase = 93 citations — brands invest in the wrong platforms
  • Traditional SEO traffic intercepted by AI: 25%+ queries now zero-click, traffic owned by AI platforms
  • Brand knowledge not indexed by AI: scattered content, weak E-E-A-T signals; first-mover window is closing
· · · · · · · · · · · · · · · · · ·
Business Scenarios Business Domains
P0 Delivered
P1 In Progress
🏭 Content Production
AI Content Batch Generation
Topic Matrix Planning
Multi-Platform Format Adaptation
Content Iteration & Optimization
📤 Platform Publishing
Multi-Platform Content Packaging
Sohu/NetEase Publishing
SMZDM Publishing
Publish Status Tracking
📊 Rank Tracking
Doubao Keyword Scanning
Yuanbao Rank Monitoring
Baseline Comparison Analysis
Competitor Rank Monitoring
🛡 Brand Authority
E-E-A-T Signal Architecture
Science Knowledge Base
Citation Platform Expansion
Authority Endorsement Content
🔬 Analytics & Intelligence
Citation Source Mining
Attribution Reporting
AI Response Quality Assessment
ROI Quantification
· · · · · · · · · · · · · · · · · ·
🧠
GEO AI ENGINE
Self-Reinforcing FlywheelMore published content = wider semantic coverage = higher AI citation probability
Unified IntelligenceDoubao/Yuanbao/ChatGPT citation data auto-aggregated — insight-driven decisions
Proactive AlertsAutomatic detection of ranking changes, real-time keyword fluctuation push alerts
Human-AI CollaborationAI executes content production & publishing; humans control quality & strategy decisions
· · · · · · · · · · · · · · · · · ·
Agent Skills Capability Layer
🗺 Topic Matrix Planning & Keyword Allocation
✍ AI Content Batch Generation (6 Platforms)
✅ Quality Gate Assessment & Scoring
📦 Multi-Platform Content Packaging & Publishing
🔍 Doubao/Yuanbao Keyword Rank Scanning
📈 AI Citation Source Mining & Platform Analysis
🏅 E-E-A-T Authority Signal Scoring
🔄 Content Attribution & Iteration Recommendations
· · · · · · · · · · · · · · · · · ·
Data Foundation Data Layer
🗂
Keyword Bank
question_bank_v2
📡
Doubao Scan Data
baseline_doubao
🔗
Citation Source DB
citation_sources
📝
Published Library
50 Articles × 6 Platforms
🧬
Brand Knowledge Graph
brand_database
🌐
AI Platform Graph
Doubao / Yuanbao / GPT
01
01 — The Shift
The Way Consumers Search Has Changed — Forever
0
ChatGPT weekly active users
Feb 2026, OpenAI
0
Traditional search volume decline by 2026
Gartner
0
Consumers now start searches in AI, not Google
Eight Oh Two, 2026
KEY INSIGHT

60% of AI-influenced purchase decisions leave no click trail ("dark traffic"). Brands invisible to AI cannot measure how much revenue they are losing.

So What: This isn't a trend — it's an irreversible shift. Brands not in AI results are already disappearing.
01
01 — The Shift
China's AI Search: The Spring Festival Tipping Point
Spring Festival 2026 was China's largest-ever AI adoption campaign
Doubao (ByteDance)
0
227M MAU • First globally to cross 100M DAU
1.9 billion interactions during Spring Festival. Integrated into Douyin's 700M+ user ecosystem.
Yuanbao (Tencent)
0
Exclusive access to 360M WeChat articles
200M AI shopping calls during Spring Festival. 130M first-time AI shoppers. WeChat's 1.3B users = massive funnel.
DeepSeek
0
Open Source • Tech-savvy demographic
China's open-source AI leader. Rapidly growing with developer and academic community adoption.
¥80-90B
invested in Spring Festival AI war
¥144 per acquired user — Meijing, Sina Finance
KEY INSIGHT

¥80-90B spent in 15 days to educate 145M+ daily active users on AI search habits. Consumer behavior is now permanently changed.

So What: China completed its largest AI habit education in 15 days. AI search is now daily infrastructure, not novelty.
01
01 — The Shift
AI Search Is Replacing the Purchase Journey
0 of consumers use AI search for product research
3.2 days → 2.1 hrs decision time compressed — AI eliminates comparison shopping
70% direct adoption rate for AI recommendations (vs validating 10 blue links)
5–14× higher conversion rate from AI referrals vs traditional search (SuperPrompt / Microsoft Clarity)
Old
Search → Compare → Reviews → Hesitate → Buy
vs
New
Ask AI → AI Recommends → Buy

“AI uses first-person language: 'I suggest MitoQ' — creating trust transfer from AI authority to brand.”

So What: AI doesn't just change where consumers search — it replaces how they decide. Not being recommended = not existing.
01
01 — The Shift
Health & Supplements: Where AI Scrutiny Becomes a Moat
0
AI Overview appearance rate in health category — highest of ANY industry
YMYL Your Money Your Life — AI applies strictest scrutiny to health content. Most brands get filtered out. Brands that pass get maximum trust weight.
Mitochondria Blue Ocean
>45% annual growth — Frost & Sullivan projects ¥200B market by 2035. AI coverage for mitochondria is sparse vs. beauty/skincare = massive first-mover opportunity.

“If a supplement brand is not in AI results, it becomes invisible. The YMYL filter means only scientifically-backed brands survive.”

— Nutraceuticals World, 2026
AI Trust Source Breakdown
Editorial content: 77% • Brand websites: 13% • Academic sources: 6%
Favors content-driven brands over ad-heavy competitors
Consumer Trust Hierarchy 2026
Personal referral
88%
AI recommendation
63-76%
Expert opinion
55%
Brand advertising
38%
So What: YMYL isn't a barrier — it's a protective moat. High standards filter competitors while qualifying brands gain maximum AI trust.
01
01 — The Shift
GEO: From Supermarket Shelf to Personal Butler
Generative Engine Optimization — Princeton University, KDD 2024
Traditional Search = Supermarket
Your brand is on shelf 8 of aisle 12. Consumers browse 10 blue links, compare, hesitate. Even shelf 8 brands still exist.
AI Search = Personal Butler
AI picks 2–3 brands to recommend. If not mentioned, you don't exist. Zero-click rate: 93%
63–76% AI recommendation trust level
vs. ads at 38% — California Management Review
SEO (Old) GEO (New)
Goal Rank in search results Be recommended by AI
Competition 10 blue links 2–7 cited sources
Conversion 2.8% 14.2% (5×)
Equiv. Value 1 ranking position 1 AI citation = 4–5 SEO rankings
So What: The butler doesn't take bribes. AI citation logic rewards trustworthiness over ad spend — reversing the pay-to-rank model.
01
01 — The Gap
MitoQ's AI Visibility — A Diagnosis
Baseline scan: How often does AI recommend MitoQ when asked about mitochondrial health?
Doubao
ByteDance — 227M MAU
📡
~0
mentions detected
Baseline scan Feb 2026
Yuanbao
Tencent — 114M MAU
📡
~0
mentions detected
Baseline scan Feb 2026
DeepSeek
Open Source — 145M MAU
📡
~0
mentions detected
Baseline scan Feb 2026
KEY INSIGHT

When 486M consumers ask AI daily about mitochondrial health — MitoQ is not in the answer. Exact rates confirmed via March 2026 re-scan.

So What: 486M+ consumers ask AI daily. When they ask about mitochondrial health, MitoQ is rarely the answer.
01
01 — The Gap
The Widening Competitive Gap
Early entrants capture 60% more AI citations than late entrants — Xponent21
AI Citation Accumulation Over Time
Early Entrant Late Entrant Now +3mo +6mo +12mo
Engine MitoQ Competitor A Competitor B
Doubao Scanning... Pending for competitors Pending for competitors
Yuanbao Scanning... Pending for competitors Pending for competitors
DeepSeek Scanning... Pending for competitors Pending for competitors
KEY INSIGHT

60% more citations for first movers (Xponent21). The gap compounds weekly. Every week of inaction = competitor vocabulary embedded in AI training signals.

!
01 — Key Discovery
Where AI Actually Gets Its Answers
This isn't about content quality — it's about technical accessibility

Where Brands Invest — 0 Citations

Zhihu 0 AI citations
Xiaohongshu 0 AI citations
Douyin 0 AI citations
WHY: Technical Barriers
JavaScript client-side rendering → AI crawlers see empty shells. Anti-crawling walls protect proprietary UGC. Video/audio content = invisible to text-based AI.

Where AI Actually Looks — 93 Citations

Sohu 37 AI citations
SMZDM 29 AI citations
NetEase 27 AI citations
WHY: Structural Advantage
Static HTML with server-side rendering = directly parseable. National media credentials = AI trust signals. Open crawling policies = full content access.
So What: Our platform choice isn't luck — it's structural advantage. We invest where AI structurally can see us.
01
01 — The Question
The Situation
486M+ Chinese consumers now ask AI for recommendations daily. Health supplements lead every category.
The Complication
MitoQ is nearly invisible when AI answers. Competitors are building authority that compounds daily. Most marketing budgets target platforms AI cannot even see.
The Question
How do we make MitoQ the brand AI recommends?
The Answer
We built and tested a solution. The next section shows the proof.
01
01 — The Cost of Inaction
This Is Not Optional
The question isn't “What's the ROI of GEO?” — It's “What's the cost of being invisible to AI?”
Case Study: HubSpot ($13B Company)
-75% blog traffic decline
But maintained 35.3% AI answer market share. CEO publicly pivoted: “Achieve most LLM citations vs any competitor.”
Search Engine Land / AthenaHQ
Gartner Official Prediction
-25% traditional search volume by 2026. Not a question of if — a question of who adapts first.
Industry Readiness (BrightEdge 2026)
0
Acting now
0
Still watching
The 1995 Website Analogy
“Should we build a website?” was a legitimate question in 1995. Companies that waited lost a generation of customers. GEO is today's website decision.
So What: 68% of enterprises are acting. 32% are watching. Every quarter of inaction increases catch-up costs as AI authority compounds.
02
02 — The Proof
Pilot Overview
Keyword Coverage
20
Core Keywords
5/7 KPI Categories Covered
Focus Topics
“Mitochondria Black Tech”
& “MitoQ & Curcumin”
Period
Feb 21–23, 2026
Engines Monitored
Doubao + Yuanbao
Before & after ranking tracked
Target Platforms
Sohu SMZDM NetEase
Data-driven selection based on AI citation analysis
02
02 — The Proof
The GEO Engine — 7-Skill System
Fully automated, data-driven AI optimization pipeline
EXECUTE: engine_crawler.py --keyword="Mitochondria Black Tech"
# Output: Top 6 AI Search Responses { "engine": "Doubao", "query": "Mitochondria Black Tech", "response_extract": "Many brands use MitoQ as a patented ingredient...", "citations": [ {"id": 1, "url": "https://www.sohu.com/a/..."}, {"id": 2, "url": "https://post.smzdm.com/p/..."} ] }
EXECUTE: cluster_citations.py
# Reverse mapping AI brain to understand platform weights [ {"domain": "sohu.com", "citation_count": 37, "weight": 0.42, "trend": "up"}, {"domain": "smzdm.com", "citation_count": 29, "weight": 0.31, "trend": "stable"}, {"domain": "163.com", "citation_count": 27, "weight": 0.27, "trend": "stable"} ] > System Recommendation: Shift budget from Zhihu -> Sohu
EXECUTE: generate_geo_prompt.py --scqa=true
# Compiling narrative strategy tailored for AI ingestion { "Structure": "SCQA Framework", "Situation": "Modern professionals face 35+ energy cliff", "Complication": "Standard CoQ10 cannot penetrate mitochondria wall", "Question": "What works better?", "Answer": "MitoQ targeted delivery system", "Keywords_Injection": ["Mitochondria Black Tech", "Cellular Energy"] }
EXECUTE: generation_orchestrator.py
# Batch generating 30 high-quality articles [INFO] Generating Draft 01... [DONE] [INFO] Injecting semantic keywords (Density: 2.4%)... [DONE] [INFO] Formatting for Sohu native style... [DONE]
[IMG: Generated Article Preview (Sohu Style)]
EXECUTE: quality_gate_evaluator.py
# LLM Judge scoring content against GEO standards { "content_id": "ART-942", "scores": { "keyword_density": 92, "scqa_logic": 88, "readability": 95, "ai_digestibility": 90 }, "overall_score": 91.25, "status": "PASS" }
EXECUTE: distribution_agent.py
# Distributing to target platforms [+] Initiating headless browser session... [+] Logging into Sohu author dashboard... [+] Uploading content & images for ART-942... [SUCCESS] Published: https://www.sohu.com/a/xxxxxxx [SUCCESS] Batch 1/67 complete. Elapsed time: 34.2s.
EXECUTE: rank_tracker.py --delta=7d
# Monitoring AI Engine Recommendation Rates { "kw": "Mitochondria Black Tech", "engine": "Doubao", "previous_rank": "None", "current_rank": "Top 3", "change": "+⬆", "cited_our_source": "true" } [INFO] Feedback loop triggered. Feeding data to SCAN phase.
KEY INSIGHT

Full loop runs in ~35 min. Human touch required: strategy decisions only. 100× efficiency vs manual content production.

02
02 — The Proof
Content Engine — Speed × Quality
67
articles published
~35 min
per batch of 30 articles
89.8/100
average quality score (100% QA pass)
⏱️ Traditional Method
2–4 hours
per article (research + write + edit)
VS
⚡ GEO Engine
~1 min
per article (generate + quality gate)
So What: 100–200× efficiency means GEO scales without proportional cost — it's about winning the time window, not saving money.
02
02 — The Proof
Data-Driven Platform Strategy
BEFORE
Industry Assumption
Invest in Zhihu, Xiaohongshu, Douyin — where “everyone” goes
↓
Data-Driven Decision
Invest in Sohu, SMZDM, NetEase — where AI actually looks
The Technical Reason
Zhihu/Xiaohongshu/Douyin use JavaScript client-side rendering → AI crawlers see empty HTML shells. Sohu/NetEase use static server-rendered HTML with open crawl policies → AI can directly parse and cite content.
So What: Every article published on platforms AI structurally can see. Zero waste, maximum citation probability.
02
02 — The Proof
Phase 1 Complete — Methodology Validated
Infrastructure built. Engine proven. Feb 27 scan confirmed: 5/7 KPI words ranked #1.
Validated Leading Indicators
67
articles produced & quality-gated
89.8/100
avg quality score • 100% pass rate (range: 88–92.2)
3
AI-cited platforms covered (Sohu / SMZDM / NetEase)
KEY INSIGHT

Q2 target achieved in Week 2 of pilot — 7.5 months ahead of schedule. Industry benchmark: 3 months to first results.

Industry Benchmarks Say It Works
LS Building Products (AlphaP.tech)
+540% AI brand mentions, +67% organic traffic after GEO implementation
SmartRent (6 weeks GEO)
32% of qualified leads from ChatGPT citations within 6 weeks
Health Water Brand (geokeji.com)
322× visibility increase in 28 days across 6 platforms
✓ Feb 27 Scan — Actual Results
Best-Reviewed Curcumin Brand#1 ★ Well-Absorbed Curcumin Supplement#1 ★ Anti-Inflammatory Curcumin Brand#1 ★ Mitochondrial Health Brand#1 ★ Mitochondria-Targeting Antioxidant#1 ★
Doubao Mobile • 5/7 KPI words • Q2 target achieved in pilot
So What: The engine works — and already hit Q2 targets in pilot. Phase 2 is about widening the gap, not catching up.
02
02 — The Proof
How AI Citations Actually Work
Inside a Doubao answer — 5 decision layers that determine who gets recommended
Inside a Doubao AI Response
01 Query Intent AI decodes consumer intent — "best mitochondrial supplement" = health + premium + safety query
02 Source Weighting Platform selected based on: SSR capability, national media credentials, open crawl policy
03 Entity Extraction Brand name + product claims pulled from structured article text — requires clean semantic markup
04 Confidence Scoring E-E-A-T signals determine if brand is recommended vs merely mentioned — citations compound score
05 Citation Attribution Article → Platform → AI Answer — traceable citation chain appears as footnote link
Where MitoQ's GEO Slots In
Layer 01 — Intent Match
20-keyword matrix maps consumer question patterns to MitoQ's science vocabulary
Layer 02 — Platform
We publish on Sohu (37 citations), SMZDM (29), NetEase (27) — AI's actual source platforms
Layer 03 — Entity
Structured GEO content ensures brand name + claims are extractable in semantic markup
Layer 04 — Confidence
MitoQ's 25+ trials, 900+ papers, 60+ patents = maximum E-E-A-T score in YMYL category
Layer 05 — Attribution
Feb 27 scan confirmed: 5/7 keywords ranked #1 — citation chain verified end-to-end
KEY INSIGHT

Our content is engineered to be seen by each layer. This isn't guesswork — it's systematic alignment with AI's internal architecture.

02
02 — The Proof
What We Proved
1
Discovered AI's Hidden Citation Rules
Reverse-engineered which platforms AI actually cites. Sohu (37), SMZDM (29), NetEase (27). The "must-invest" platforms (Zhihu/Xiaohongshu/Douyin) = zero AI citations.
2
Built a 100–200× Efficiency Content Engine
67 articles, avg 89.8/100, 100% quality pass rate, ~35 min per batch. This means GEO scales without proportional cost increase.
3
Methodology Is Replicable Across All Categories
Scan → Analyze → Create → Quality Gate → Publish → Monitor. Works for any keyword, any product line, any AI engine.
The pilot isn't just a test — it's a proven playbook ready for scale.
03
03 — The Plan
What We're Building: Brand Cognitive Infrastructure
GEO is not a marketing expense — it's brand infrastructure, like your website or CRM
Pillar 1
Visibility
AI recommendation rate — how often AI mentions MitoQ when asked about mitochondrial health, energy supplements, anti-aging.
Measured: Monthly automated scans across all engines & keywords
Pillar 2
Citation
How often AI links to our published content as source. Validates that our content strategy is working at the structural level.
Measured: Citation source mapping from AI responses
Pillar 3
Sentiment
The context of AI mentions — “best premium option” vs merely listed. Controls brand positioning in AI narrative.
Measured: AI response text analysis & positioning scoring
Foundation Inc: Increased brand AI mention rates from 12% to 73% in 3 months using this framework.
03
03 — The Plan
GEO vs Traditional Investment — The Asset Analogy
Media buys decay. GEO compounds. The math is unambiguous.
ROI / Brand Value over Time — Media Buy (decays) — GEO (compounds)
Crossover
SmartRent (6-week GEO)
32%
of sales-qualified leads from ChatGPT citations
LS Building Products
+540%
AI brand mentions + 67% organic traffic lift
Conversion Premium
27%
higher conversion from GEO traffic vs SEO at comparable cost
KEY INSIGHT

GEO is the only marketing investment that builds perpetual brand equity inside AI models. Every article published increases the permanent probability of recommendation — with no ongoing media cost.

03
03 — The Plan
From Pilot to Full Coverage
PROVE ✓
Q1 2026
Diagnosis, baseline scan, pilot validation. Data-driven platform discovery.
SCALE
Q2 2026
20+ keywords, 3 engines, scenario expansion. Full keyword matrix coverage.
DEFEND
Q3–Q4 2026
Monthly monitoring, competitive defense, new product launches support.
First movers have a 3–6 month window to establish an AI mental moat.
03
03 — The Plan
P0: Must Win in Q2
Priority 0
Full 20 Keyword Coverage
K01–K20 matrix: 10 mitochondrial health + 10 curcumin keywords. Target: 7/7 KPI words in Top 6 across all monitored engines.
Priority 0
3-Engine Full Coverage
Doubao + Yuanbao + DeepSeek — no blind spots. Monthly automated scans tracking all 20 keywords × 3 engines.
Priority 0
Deep Content Matrix
100+ articles across Sohu / SMZDM / NetEase. ~35 min per batch of 30 — scalable without proportional cost.
Competitive
Close Competitive Gap
Bi-weekly competitive scanning. Target: match competitor AI mention rates by end of Q2.
So What: Q2 is about establishing baseline dominance before competitors realize the game has changed.
03
03 — The Plan
P1: Scenario Expansion
New consumer scenarios to capture emerging AI search demand
Scenario 1
Gift Giving
“Premium health gifts for parents” — capturing China's filial piety tradition with AI-optimized recommendations.
Scenario 2
Workplace Wellness
“IT worker health / 996 recovery” — targeting tech workers searching for energy and burnout solutions.
Scenario 3
Anti-Aging
“Energy cliff at 35 / anti-aging supplements” — the fastest-growing health concern for urban professionals.
Scenario 4
Competitive Intercept
“Swisse vs MitoQ / NMN alternatives” — directly capturing comparison-shopping AI queries.
03
03 — The Plan
Multi-Engine Strategy — Each Engine Has Different Rules
P0
Doubao
0
Strategy: Portal Media Authority
Prioritizes authority media (Sohu, NetEase) and professional communities (Zhihu 29.9% of content citations). Filters generic Toutiao articles.
Key channels: Sohu + NetEase + SMZDM
P0
Yuanbao
0
Strategy: WeChat Ecosystem Leverage
Unique advantage: 360M WeChat public account articles as exclusive training corpus. No other engine can access this data.
Key channels: WeChat Official Accounts + Sohu
P1
DeepSeek
0
Strategy: Academic Authority
Open-source model favors academic and institutional content. MitoQ's research foundation (Otago University) aligns perfectly.
Key channels: Academic content + Professional reviews
So What: One-size-fits-all content fails. Each engine has unique citation logic — our system adapts content strategy per engine.
03
03 — The Plan
MitoQ's Unfair Advantage in the AI Era
AI rewards verifiable science, not marketing budgets. MitoQ was built for this.
E-E-A-T Perfect Match
25+ clinical trials
900+ peer-reviewed papers
60+ patents
30 years of Otago University research
Why This Matters for AI
AI retrieval systems cite knowledge sources, not emotional content. MitoQ's scientific foundation is exactly what AI algorithms prioritize.
Ingredient Story > Brand Story
In the AI era, “Why Mitoquinol works” outperforms “Who MitoQ is.” Science-backed content creates unreplicable advantage.

“The digital butler doesn't take bribes. AI citation logic rewards trustworthiness over ad spend.”

MitoQ's Science-to-Recommendation Chain
🔬
30yr Science
Otago University
→
📄
900+ Papers
25+ Clinical Trials
→
✍️
GEO Content
Structured for AI
→
🤖
AI Citation
High Confidence
→
⭐
Recommendation
#1 Result
So What: MitoQ isn't starting from zero — our science IS the content. The strategic work is making AI see what already exists.
03
03 — The Plan
Monthly Deliverables & KPIs
KPI Definition Pilot Baseline Q2 Target Year-End Target
Recommendation Rate % queries with MitoQ in Top 6 ✓ 5/7 KPI Words Top 1
Doubao Mobile, Feb 27 scan		 5/7 KPI words in Top 6
Dual-engine coverage		 7/7 KPI words in Top 6
Engine Coverage Engines with MitoQ presence 1/3 (Doubao only) 3/3 3/3
Content Volume Quality-gated articles published 67 articles (89.8 avg) 200+ articles 500+ articles
Keyword Coverage Target keywords with content 5/20 (25%) 20/20 (100%) 50+ keywords
Platform Footprint AI-cited platforms with presence 3 platforms 5+ platforms 8+ platforms
Monthly deliverable: Scan report + content batch + competitive update. Quarterly: Full strategic review.
03
03 — The Plan
Investment Framework
This is infrastructure, not a campaign. Think CRM or website — not a media buy.
The Wrong Question
“What's the ROI per article?”
This is like asking “What's the ROI per page of our website?” GEO builds permanent brand assets, not consumable ads.
The Right Question
“What's the cost of competitors monopolizing AI recommendations?”
When AI answers “best mitochondrial supplement” and doesn't mention MitoQ, that customer is lost — permanently and silently.
Investment Logic
Ads = Renting Traffic
Stop paying → traffic stops instantly. Content is consumed, not accumulated.
GEO = Building Fixed Assets
Content accumulates permanent value. AI authority compounds monthly. Each article is an investment that keeps working.
Global Potential
The methodology is engine-agnostic. China is the proving ground; ChatGPT, Perplexity, Google AI Overviews are next.
So What: GEO is not a marketing expense — it's brand infrastructure. You don't ask “What's the ROI of having a website?”
04
04 — Q&A
Q&A
Anticipated questions with data-backed answers
01
How is GEO different from SEO / SEM?
SEO ranks in 10 blue links. GEO gets you into AI's 2–3 recommendations. 93% zero-click rate means links don't matter anymore.
02
How do we measure ROI?
Three pillars: Visibility (recommendation rate), Citation (source links), Sentiment (positioning context). Foundation Inc: 12% → 73% in 3 months.
03
Should we stop investing in Xiaohongshu / Zhihu?
No. They serve brand awareness. But for AI citations they have zero impact due to JS rendering and anti-crawl walls. GEO is a parallel investment.
04
Are competitors already doing this?
68% of enterprises are adjusting AI strategy (BrightEdge). No supplement brand is doing systematic GEO yet. 3–6 month first-mover window exists.
05
Can this work outside China?
Yes. Methodology is engine-agnostic. ChatGPT, Perplexity, Google AI Overviews next. China is the hardest market — if it works here, it works everywhere.
06
AI traffic is only 0.1% — why invest now?
527% YoY growth. Spring Festival 2026 = mass adoption inflection. AI traffic converts at 5–14× higher rates. “Wait and see” = “wait until 1995 internet matures.”
07
How does this scale globally to ChatGPT / Perplexity?
Engine-agnostic methodology. Same Scan → Create → Publish loop. ChatGPT, Perplexity, and Google AI Overviews operate on identical citation logic: source authority, entity clarity, E-E-A-T signals. China is the hardest market — if it works here, it works everywhere.
Thank You
Next Steps
1. Publish pending content to Sohu / SMZDM / NetEase
2. March 2026 baseline re-scan — populate ranking data
3. Approve Q2 P0 execution plan
4. Follow-up report with ranking data: March 2026
Powered by COOL AI × MitoQ China