Key facts and TL;DR
The signals and decisions leadership needs first.
U.S. MARKET INTELLIGENCE · 150-COMPANY RESEARCH COHORT
See your U.S. market through a CMO's eyes
MegaCMO connects company evidence, category positioning, website experience, search visibility, content, competitors, and demand generation into an actionable U.S. market view.
United States · English · Source-linked company research
U.S.-specific cohort · Automated multi-agent review · Unverifiable facts remain N/A
U.S. research scale
150 U.S. company reports are published
Every U.S. report passed the source, citation, template, artifact, and automated multi-agent review gates recorded in the publication manifest.
- 150
- Frozen U.S. research identities
- 2,476
- Registered source citations
- 150
- Published full reports
- 15
- U.S. industry lanes
Current state: 0 in research, 150 published. Publication state is read directly from the U.S. manifest.
U.S. cohort coverage
150 U.S. companies, one published research standard
90 benchmark leaders, 30 growth challengers, and 30 startup frontiers span 15 U.S. industry lanes. 150 are published and 0 remain in research.
These are public-data research subjects, not customers or partners. 150 U.S. reports are currently published; researching identities expose no analytical conclusions.
A complete diagnosis you can take straight into the boardroom
Nine connected chapters turn market evidence into an actionable plan.
Audience and personas
Who needs you most, and what shapes their decision.
Positioning and SWOT
Your defensible strengths, risks, and market opening.
Website and UX review
Five prioritized conversion wins to address first.
SEO and keyword gaps
Where demand exists and your visibility falls short.
Competitor landscape (value-ranked)
The whole battlefield mapped: the giants, the dark horses, and where you sit.
Content and social review
Content gaps translated into a 90-day calendar.
Paid media and funnel blueprint
A connected TOFU–MOFU–BOFU acquisition path.
KPIs and 30-60-90 plan
Priorities, owners, and measures for execution.
One URL. Three clear steps.
Share a URL
Your own website - or a competitor's. Choose a report language, pay online, and add your biggest marketing question for a more focused diagnosis.
Frontier AI research
A frontier model studies the website and public market signals. Deep-dive plans run the full research pass and write themselves, including a diagnostic annotation that is labelled as AI-written.
Receive it in 10 minutes or 24 hours
Light can arrive in as little as 10 minutes; deep-dive plans arrive within 24 hours. We email a private report link that is neither published nor indexed by default.
Analysis engine
Frontier models do the reasoning. The pipeline stops them inventing things.
Every report clears the same auditable pipeline and leaves ten audit artifacts behind. If any gate fails, the report is not published.
- 01
Source registration
Each citation carries a URL, publisher, reliability tier, and access time, spanning at least two independent domains. Sources are never shared across companies.
- 02
Claim binding
Each material fact binds to between one and four sources. A statement without a source cannot be written as fact, only as clearly labelled analysis.
- 03
Independent review
A separate pass confirms each source is genuinely reachable, that publisher and title match, and that no figure has been inflated. Whatever it finds gets corrected.
- 04
Contradictions and gaps
Where sources conflict or nothing verifiable exists, the report says N/A and states how far verification went. No estimates fill the gap.
- 05
Bilingual parity
Traditional Chinese and English are aligned section by section, each written for its own reader rather than machine translated.
- 06
Static packaging
Reports ship with no scripts, no external assets, and no tracking, so they read offline and archive cleanly, with the version locked by hash.
Rules we enforce on ourselves
- Traffic, domain authority, keyword volume, CPC, and ad spend are marked N/A without a public source. We do not estimate them.
- Market-ranking, financial, medical, and regulatory claims require extra review, and medical or aesthetic services never get efficacy or outcome claims.
- Every report states its own review status in the body. An AI review is never presented as a human fact-check.
- Reports are private and non-indexable by default, and paid reports carry no sales links.
Why not just ask ChatGPT?
You can. The difference is:
A fixed methodology
Nine chapters and four required comparison tables - not a one-off conversation.
Checked before it ships
Every deep report clears a fixed set of automated safety checks before it is sent: no sales links, no external assets or scripts, privacy headers in place, quality signature intact. Fail any one of them and it is not delivered.
A decision-ready artifact
A polished report you can share with your team or board, not a transcript you still need to interpret.
U.S. shipping log
A market-specific research system
The U.S. cohort moves through source collection, analysis, independent automated review, and content-addressed release.
U.S. market foundation
150 company identities frozen across 15 industry lanes with separate en-US routes and catalogue ownership.
Dynamic research workflow
Bounded source collection and independent fact, compliance, editorial, and QA agents run with resumable checkpoints.
Six-company pilot
Microsoft, CoreWeave, OpenAI, Walmart, Warby Parker, and Whatnot established the first evidence and template contract.
Single visual standard
Every U.S. report uses the same TSMC-derived MegaCMO report shell; agents fill content without inventing layouts.
PRICING
One business dinner for a year of marketing direction
Light Rapid Diagnosis
A concise single-language report generated by Claude Fable 5
NT$69910 MIN · AI ONLY
- As fast as 10 minutes
- Concise diagnosis + mini landscape (8-12 companies)
- Generated by Claude Fable 5
- Great for quick scans and competitor recon
- Full Light credit toward an upgrade
Deep Diagnosis: Full Battlefield
Complete diagnosis + a 30-50 company landscape + close-quarters rival deep-dives
NT$5,490USUALLY 1 HR · FULLY AUTOMATED
- Complete frontier-model diagnosis
- 30-50 company landscape (value-ranked, with giants/dark-horses/direct-rival reads)
- Two direct-rival deep-dives
- AI diagnostic annotations, labelled in the report as AI-written
- Generated end to end by AI, usually within an hour and always within 24 hours
Full Battlefield + Consultant Debrief
Everything in Full Battlefield plus a 30-minute one-to-one debrief
NT$19,900
- Everything in Full Battlefield: the same report as the NT$5,490 plan
- A 30-minute one-to-one debrief with a real consultant
- The full report reaches you before the session, so you can bring questions
- Report usually within an hour, always within 24 hours
Extended service · We arrange the 30-minute session with you separately
Choose this planApply your full Light payment toward an upgrade · NT$500 off public deep-dive reports
Powered by Claude Fable 5 & GPT-5.6 · Charged in TWD
What to know before you order
01What data do you use?
Only public information: your website, search results, public social channels, and market data. Anything unavailable is marked N/A - never invented.
02Which AI models run the analysis?
Frontier models: Claude Fable 5 and GPT-5.6. Light is generated end-to-end by Fable 5; deep-dive plans run the full research pass and are also generated end to end, clearing a fixed set of safety checks before delivery. No one rewrites them by hand.
03Can I analyze a competitor, or a business I haven't started yet?
Yes - and every report includes a competitor landscape by default. Any public website works: your own, a competitor's, or a benchmark brand. Aspiring founders use it to scan a target market; for the full 30-50 company battlefield, choose a deep-dive plan.
04Will my report be public?
No. Reports use a private random link and are not indexed. We only publish a deep-dive report when you explicitly opt in, which also gives you NT$500 off. Light does not offer the public-report discount.
05When will I receive it?
Light arrives within 10 minutes of payment; deep-dive plans arrive within 24 hours.
06How is Light different from a deep-dive plan?
Light uses Claude Fable 5 for a concise single-language report including a mini landscape of 8-12 competitors. Deep-dive plans run the complete diagnosis plus a full 30-50 company value-ranked landscape with direct-rival deep-dives, generated end to end by AI, with a diagnostic annotation that is labelled as AI-written.
07Can I choose the language?
Light is available in either Traditional Chinese or English. Deep-dive plans can also include a second-language edition.
08Can I get a refund?
No. Reports are custom digital content and production begins immediately after payment, so all orders are non-refundable. If a delivered report has a material quality defect, contact us within 7 days for a free re-run or account credit.
U.S. research cohort
Start with a published U.S. company report
All 150 U.S. company reports are published with source-linked previews and entitlement-controlled full artifacts.
Microsoft
Published public-data preview. Full report access remains entitlement-controlled.
View previewUnitedHealth Group
Published public-data preview. Full report access remains entitlement-controlled.
View previewTesla
Published public-data preview. Full report access remains entitlement-controlled.
View previewNT$699. Ten minutes. A fresh CMO perspective.
One report aligns your brand, search, content, and growth priorities.