feat(Notion): Refactor product DBs for GTM strategy

- Implements a 3-tier database architecture (Canonical Products, Portfolio, Companies) to separate product master data from company-specific portfolio information.

- Upgrades import_competitive_radar.py to an intelligent "upsert" script that prevents duplicates by checking for existing entries before importing.

- This enables detailed GTM strategy tracking for RoboPlanet products while monitoring competitor portfolios.

- Updates documentation to reflect the new architecture and import process.

MIGRATION_REPORT_COMPETITOR_ANALYSIS.md

@@ -148,6 +148,44 @@ Neben der reinen Analyse wurde das Fundament für ein dauerhaftes Monitoring-Sys
 3.  **Relationaler Import (v6):**
     *   Der Notion-Import wurde auf **v6** aktualisiert. Er unterstützt nun lückenlos die neue v5-Struktur, importiert Chain-of-Thought Beschreibungen in Rich-Text-Felder und verknüpft erstmals auch die extrahierten **Referenzkunden** relational in Notion.
 
+
+### 📡 Post-Migration Architectural Refactoring (Jan 22, 2026): From Unified Import to GTM-Ready Architecture
+
+**Problem Statement:** The v6 import structure, while relational, suffered from a fundamental design flaw: the informational depth of a product entry was identical for all vendors. A product sold by RoboPlanet was captured with the same superficial data fields as a competitor's product. This prevented the mapping of our detailed Go-to-Market strategies and, due to the "Purpose" texts generated by the LLM, continued to create duplicates, as identical products received slightly different descriptions.
+
+**Strategic Decision: Separation of "What" from "How"**
+
+To create a true "Single Source of Truth" and to enrich our own products with the necessary depth (GTM strategy, support knowledge, etc.), a new architecture was decided upon. The system is being converted from a 2-tier model (Companies ↔ Products) to a 3-tier model.
+
+**The New 3-Tier Architecture:**
+
+1.  **🆕 Canonical Products (The "WHAT" Database):**
+    *   A completely new database that contains each product model (e.g., "Puma M20") **only once**.
+    *   Content: Objective master data (manufacturer, model, technical specs) and a relation to the `Product Categories` DB.
+    *   This is the market-wide, vendor-neutral truth.
+
+2.  ** repurposed Portfolio (The "HOW" Database):**
+    *   The existing `Competitive Radar (Products)` database is being repurposed and renamed.
+    *   It functions as a **junction table**. Each entry represents the relationship between a vendor and a canonical product.
+    *   Example Entries:
+        *   "RoboPlanet sells Puma M20"
+        *   "TCO Robotics sells Puma M20"
+    *   **Crucially:** This database contains the **context-specific information**. For RoboPlanet entries, the fields for the complete GTM strategy (target industries, pain points, battle cards, ROI logic, etc.) will be stored here. For competitor entries, these fields will remain empty.
+
+3.  ** Companies (The "WHO" Database):**
+    *   Remains unchanged and contains the master data of the competitors and of RoboPlanet itself.
+
+**New Relational Link:**
+`[Canonical Products]` ↔️ `[Portfolio]` ↔️ `[Companies]`
+
+**Migration Plan:**
+A one-time Python script will perform the "Operation Clean Architecture":
+1.  **Schema Transformation:** Creation of the `Canonical Products` DB, conversion of the old product DB to the `Portfolio` DB.
+2.  **Intelligent Migration:** Reading of the old entries, creation of the unique entries in `Canonical Products`, and subsequent re-linking of the (now) `Portfolio` entries.
+3.  **Categorization:** Assignment of the canonical products to the global `Product Categories`.
+
+**Result:** This refactoring elevates the system from a pure market intelligence tool to a true **Strategic Marketing OS** that can directly map and support our own sales and marketing processes.
+
 ---
 *Dokumentation finalisiert am 12.01.2026.*