58 lines
2.0 KiB
Markdown
58 lines
2.0 KiB
Markdown
# Competitive Landscape Research Pipeline
|
|
|
|
Systematic competitive analysis pipeline — scrape target sites, extract features, cross-reference against ITPP products, and identify new product opportunities. Designed as a repeatable research operation.
|
|
|
|
**Status:** LIVE
|
|
**Owner:** Germaine Brown / Sho'Nuff
|
|
**Created:** 2026-08-10
|
|
|
|
## Access
|
|
|
|
| Resource | URL | Access |
|
|
|---|---|---|
|
|
| Git Repo | https://git.itpropartner.com/ippadmin/competitive-landscape-research | ippadmin (read/write) |
|
|
| AI Costs | admin-ai key: `competitive-research-prod` | TBD |
|
|
|
|
## Purpose
|
|
|
|
When Germaine identifies a batch of competitor/adjacent sites, this pipeline:
|
|
1. Scrapes and deeply analyzes every site
|
|
2. Extracts features, UI patterns, pricing models, and UX decisions
|
|
3. Cross-references findings against ITPP products (IntelSight, HotNow, VerdictTank, TransitPin, VentureBuilt, DRE)
|
|
4. Identifies new product opportunities from common themes
|
|
5. Produces a structured, searchable report
|
|
|
|
## Pipeline Architecture
|
|
|
|
```
|
|
Research Batch (user provides URLs)
|
|
│
|
|
├─ Team Alpha (3-7 sites) ──┐
|
|
├─ Team Bravo (3-7 sites) ──┤ Parallel scraping + analysis
|
|
└─ Team Charlie (3-7 sites)─┘
|
|
│
|
|
▼
|
|
Integration & Synthesis (master report)
|
|
│
|
|
├─ Feature Extraction Matrix
|
|
├─ Product Enhancement Recommendations
|
|
├─ New Product Opportunities
|
|
└─ Risk Flags & Cost Summary
|
|
```
|
|
|
|
## Research Batches
|
|
|
|
| Batch ID | Date | Sites | Researchers | Products Analyzed | Report |
|
|
|---|---|---|---|---|---|
|
|
| 001 | 2026-08-10 | 20 sites | Alpha, Bravo, Charlie | 6 products | research/001-aug10.md |
|
|
|
|
## Methodology
|
|
|
|
See [METHODOLOGY.md](METHODOLOGY.md) for the full research protocol.
|
|
|
|
Key principles:
|
|
- Every site gets a full page extraction + browser inspection
|
|
- Features are rated on replicability and mapped to specific ITPP products
|
|
- New product ideas must cite which sites inspired them
|
|
- AI costs tracked per-research-batch via dedicated LiteLLM key
|