Sandeep Lamba
TL;DR: Used Claude's research mode to generate a 15-page IFE market analysis, then had Claude build a full Next.js website to visualize it—all in under an hour.
Market research generates dense reports. My Claude research query on in-flight entertainment systems returned 15 pages of structured analysis covering oligopolies, financial metrics, competitive matrices, and regulatory frameworks. Valuable insights buried in walls of text.
Reading time: 2+ hours. Attention span: 15 minutes.
Instead of manually building a dashboard, I meta-programmed the process:
Input: Market structure analysis prompt for IFE systems
Mode: Claude Web + Research
Output: Claude.pdf (15 pages, investor-grade dossier)
Key requirement: The prompt enforced strict sourcing rules—only primary filings (10-K, 20-F), regulatory documents, and verified tier-one research. Every metric traceable or tagged "single-source (low confidence)."
The prompt is like below. Please change the sector name to your sector name.
My sector name - In-flight entertainment (IFE) systems for commercial aviation
## ROLE
Market-structure analyst for long-term investors. Produce an investor-grade, qualitative-first dossier.
Facts must be traceable and cross-verified. If evidence is weak, say so.
## INPUTS
sector = YOUR_SECTOR_NAME
## OBJECTIVE
Understand
- how the sector is structured
- who the true players are, and
- how competition works.
Deliver a 1-page executive overview + a full structured analysis with compact tables.
## ENFORCEMENT RULES (NO GUESSING)
Source bounds:
- use only A–D.
- If none available → write “Insufficient evidence.”
A) Primary filings/IR (10-K/20-F/URD/S-1, audited IFRS/GAAP, investor days, transcripts).
B) Regulators/statistics (industry regulators, national stats, IMF/OECD, BIS, etc.).
C) Official price lists, methodologies, technical specs, product catalogs, rulebooks.
D) Tier-one journals/consultancies with disclosed primary data; major data vendors with methodology.
### Verification
Any figure must be checked against ≥2 independent A–C sources → else tag “single-source (low confidence).”
### Uncertainty
If unsure → write “Unknown.” Do not estimate.
### Conflicts
If sources differ >5% → show both, define scope/period, and explain likely cause.
### Recency
Prefer ≤24 months; older = “Legacy.”
### Citations
Bracketed footnotes [1], [2], mapping to a Source List (≤15 items).
## UNIT OF COMPETITION (MANDATORY) :
For the given sector, automatically identify:
1. Lanes = the 5 universal types of competition
(Only use those that exist in the sector.)
- Upstream Inputs / Extraction
- Core Production / Infrastructure / Operations
- Distribution / Access / Channels
- Data / IP / Standards / Certification
- Services / Wrap-around Support
2. Customer groups = the 5 universal buyer types
(Only use those that exist in the sector.)
- End consumers
- Enterprise/professional buyers
- Intermediaries/distributors
- Regulators/standards bodies
- Complementors/ecosystem partners
3. Define true peers per lane
- Do not compare players across unrelated lanes.
## OUTPUTS
A) Executive Overview (≤300 words)
- Sector structure, lanes, main players, differentiation, constraints, catalysts.
B) Player Atlas (qualitative)
- For every significant player in each lane:
- Legal/brand; ownership; geography in North America, Europe, Asia-Pacific
- Vertical scope (upstream → core → distribution → data/IP → services).
- Products/offerings.
- Customers served.
- Why customers choose them (price, quality, reliability, service, design, IP, network, availability, latency, brand, regulation, etc.).
- Pricing model.
- Market-design specifics (industry-dependent).
- Competitive stance (local monopoly, fragmented, oligopoly, global scale, niche).
- Moats: network effects, IP, scale, regulation, switching costs, ecosystem lock-in.
- Vulnerabilities: regulation, commoditization, disruption, outages, supply risk.
C) Peer Sheet — Normalized KPIs (per lane)
- Revenue/EBIT mix.
- Margin structure (EBITDA %, FCF conversion).
- ROCE/ROIC.
- Capex intensity.
- Pricing power indicators.
- Volume metrics relevant to the sector.
- Productivity/efficiency metrics.
- 10-year through-cycle resilience.
D) Moats & Switching Analysis
- Lane-by-lane qualitative assessment.
E) Customer-Choice Narratives
- For each customer group: why they choose one supplier over another.
F) Regime, Regulation, & Catalysts
- Regulatory constraints.
- Structural changes.
- M&A patterns (10-year).
- Technology shifts.
- Macro drivers and risks.
G) Watchlist (5–8 items)
Key KPIs, catalysts, and risks to monitor.
## COMPARATIVE MATRICES (MANDATORY)
1) Players × Customer-Choice Drivers
(price, quality, service, network, distribution, IP, regulation, reliability, etc.)
2) Players × Moats/Frictions
(network effects, IP, scale economics, regulatory entrenchment, switching costs)
3) Lane-based cluster map
Shows true peers vs adjacent players.
## FORMAT
Do the following:
- 1-page Executive Overview
- Full Player Atlas & Peer Sheet
- Two matrices + short narratives
- Appendices with
- Assumptions
- Methods, and
- CSV-ready tables
## BEGIN
Do the following:
- Identify the lanes (from the universal categories).
- Map players into lanes; define true peers.
- Build Player Atlas.
- Populate Peer Sheet using evidence rules.
- Write customer-choice narratives + moat analysis.
- Add regime, catalysts, and watchlist.
- Deliver all sections.
Asked Claude to write deployment instructions for building a website from the PDF. The generated instructions.md specified:
Dropped instructions.md into a new Cursor/Claude session. The AI generated:
Zero manual coding. Just iterative refinements through prompts.
Cursor Project looks like this
Tables use semantic colors automatically:
● (strong) → bg-green-100 text-green-800◐ (moderate) → bg-yellow-100 text-yellow-800○ (weak) → bg-gray-100 text-gray-600The site preserves research rigor:
Complex financial tables adapt responsively:
Traditional workflow:
| Phase | Traditional Workflow | Time | AI-Assisted Workflow | Time |
|---|---|---|---|---|
| Research | Commission research | 2 weeks | Claude research | 12 minutes |
| Analysis | Analyst reads/synthesizes | 8 hours | Generate site specs | 5 minutes |
| Design | Designer creates deck | 16 hours | — | — |
| Development | Developer builds dashboard | 40+ hours | Claude builds site | 40 minutes |
| TOTAL | ~80 person-hours | ~1 hour | ||
| ROI | ~64x time multiplier |
The deployed site includes:
Run locally:
bash
npm install && npm run dev
This approach works across multiple domains:
Business Intelligence:
Knowledge Management:
The research prompt is sector-agnostic. Simply modify the input:
sector = [Your Industry]
Examples:
The instruction template automatically adapts to your domain, generating appropriate data structures, visualizations, and competitive frameworks.
The future of technical work transcends code volume. It demands exceptional specification writing, intuitive information architecture, and the ability to distill complex data into clear visual narratives.
More on SDD(Spec Driven Design) in future blogs
No comments yet. Be the first to comment!