feat(notion): Implement relational Competitive Radar import

- Added import_relational_radar.py for bidirectional database structure in Notion.
- Added refresh_references.py to populate analysis data with grounded facts via scraping.
- Updated documentation for Competitive Radar v2.0.
This commit is contained in:
2026-01-11 11:57:43 +00:00
parent 1f88b8ae20
commit e1d115e0ba
5 changed files with 482 additions and 11 deletions

View File

@@ -69,12 +69,15 @@ Die App ist unter `/ca/` voll funktionsfähig und verfügt nun über eine "Groun
* **Map-Reduce:** Statt eines Riesen-Prompts werden Konkurrenten parallel einzeln analysiert. Das skaliert linear.
* **Logging:** Ein spezieller `log_debug` Helper schreibt direkt in `/app/Log_from_docker`, um Python-Logging-Probleme zu umgehen.
### Lessons Learned für die Ewigkeit
### 📊 Relationaler Notion Import (Competitive Radar v2.0)
Um die Analyse-Ergebnisse optimal nutzbar zu machen, wurde ein bidirektionaler Import-Prozess nach Notion implementiert (`import_relational_radar.py`).
* **Architektur:** Statt Textblöcken werden drei vernetzte Datenbanken erstellt:
1. **📦 Companies (Hub):** Stammdaten, USPs, Portfolio.
2. **💣 Landmines (Satellite):** Einzelfragen und Angriffsvektoren, verknüpft mit der Company.
3. **🏆 References (Satellite):** Konkrete Kundenprojekte, verknüpft mit der Company.
* **Dual-Way Relations:** Dank `dual_property` Konfiguration sind die Verknüpfungen in Notion sofort in beide Richtungen navigierbar (z.B. sieht man auf der Company-Seite sofort alle zugehörigen Landmines).
* **Daten-Qualität:** Durch die Map-Reduce Analyse und das gezielte Reference-Scraping werden nun echte Fakten statt KI-Halluzinationen importiert.
1. **F-STRINGS SIND VERBOTEN** für Prompts und komplexe Listen-Operationen.
2. **TRIPLE RAW QUOTES (`r"""..."""`)** sind der einzige sichere Weg für Strings in Docker-Umgebungen.
3. **DUAL SDK STRATEGY:** Legacy SDK für Stabilität (`gemini-2.0-flash`), Modern SDK für Spezial-Features.
4. **MAP-REDUCE:** Bei Listen > 3 Elementen niemals das LLM bitten, "alle auf einmal" zu bearbeiten. Immer zerlegen (Map) und aggregieren (Reduce).
5. **SCHEMA FIRST:** Frontend (`types.ts`) und Backend (`Pydantic`) müssen *vorher* abgeglichen werden. `422` bedeutet fast immer Schema-Mismatch.
---
*Dokumentation aktualisiert am 11.01.2026 nach erfolgreicher Skalierung auf 9+ Konkurrenten.*
*Dokumentation aktualisiert am 11.01.2026 nach Implementierung des relationalen Competitive Radars.*

View File

@@ -39,10 +39,11 @@ Die Schaltstelle für die hyper-personalisierte Ansprache.
* **Logik:** Trennung in **Satz 1** (Individueller Hook basierend auf der aktuellen Website-Analyse des Zielkunden) und **Satz 2** (Relationaler Lösungsbaustein basierend auf Branche + Produkt).
* **Voice-Ready:** Vorbereitung von Skripten für den zukünftigen Voice-KI-Einsatz im Vertrieb und Support.
### 3.4 Competitive Radar (Market Intelligence)
Automatisierte Überwachung der Marktbegleiter.
* **Funktion:** Kontinuierliches Scraping von Wettbewerber-News und Blogposts.
* **Kill-Argumente:** Direkte Gegenüberstellung technischer Specs zur Erstellung von Battlecards für den Sales-Außendienst.
### 3.4 Competitive Radar (Market Intelligence v2.0)
Automatisierte Überwachung der Marktbegleiter mit Fokus auf "Grounded Truth".
* **Funktion:** Kontinuierliches Scraping von Wettbewerber-Webseiten, gezielte Suche nach Referenzkunden und Case Studies.
* **Kill-Argumente & Landmines:** Erstellung von strukturierten Battlecards und spezifischen "Landmine Questions" für den Sales-Außendienst.
* **Relationaler Ansatz:** Trennung in drei verknüpfte Datenbanken (Firmen, Landmines, Referenzen) für maximale Filterbarkeit und Übersicht.
### 3.5 Enrichment Factory & RevOps
Datenanreicherung der CRM-Accounts.
@@ -90,6 +91,8 @@ Um die relationale Integrität zu wahren, sind folgende Datenbanken in Notion zw
* **Product Master** $\leftrightarrow$ **Sector Master** (Welcher Roboter passt in welchen Markt?)
* **Messaging Matrix** $\leftrightarrow$ **Product Master** (Welche Lösung gehört zum Text?)
* **Messaging Matrix** $\leftrightarrow$ **Sector Master** (Welcher Schmerz gehört zu welcher Branche?)
* **Competitive Radar (Companies)** $\leftrightarrow$ **Competitive Radar (Landmines)** (Welche Angriffsfragen gehören zu welchem Wettbewerber?)
* **Competitive Radar (Companies)** $\leftrightarrow$ **Competitive Radar (References)** (Welche Kundenprojekte hat der Wettbewerber realisiert?)
* **The Brain** $\leftrightarrow$ **Product Master** (Welches Support-Wissen gehört zu welcher Hardware?)
* **GTM Workspace** $\leftrightarrow$ **Product Master** (Welche Kampagne bewirbt welches Gerät?)
* **Feature-to-Value Translator** $\leftrightarrow$ **Product Master** (Welcher Nutzen gehört zu welchem Feature?)

View File

@@ -0,0 +1,200 @@
import json
import os
import requests
import sys
# Configuration
JSON_FILE = 'analysis_robo-planet.de.json'
TOKEN_FILE = 'notion_token.txt'
# Root Page ID from notion_integration.md
PARENT_PAGE_ID = "2e088f42-8544-8024-8289-deb383da3818"
DB_TITLE = "Competitive Radar 🎯"
def load_json_data(filepath):
try:
with open(filepath, 'r') as f:
return json.load(f)
except Exception as e:
print(f"Error loading JSON: {e}")
sys.exit(1)
def load_notion_token(filepath):
try:
with open(filepath, 'r') as f:
return f.read().strip()
except Exception as e:
print(f"Error loading token: {e}")
sys.exit(1)
def create_competitor_database(token, parent_page_id):
url = "https://api.notion.com/v1/databases"
headers = {
"Authorization": f"Bearer {token}",
"Notion-Version": "2022-06-28",
"Content-Type": "application/json"
}
payload = {
"parent": {"type": "page_id", "page_id": parent_page_id},
"title": [{"type": "text", "text": {"content": DB_TITLE}}],
"properties": {
"Competitor Name": {"title": {}},
"Website": {"url": {}},
"Target Industries": {"multi_select": {}},
"USPs / Differentiators": {"rich_text": {}},
"Silver Bullet": {"rich_text": {}},
"Landmines": {"rich_text": {}},
"Strengths vs Weaknesses": {"rich_text": {}},
"Portfolio": {"rich_text": {}},
"Known References": {"rich_text": {}}
}
}
print(f"Creating database '{DB_TITLE}'...")
response = requests.post(url, headers=headers, json=payload)
if response.status_code != 200:
print(f"Error creating database: {response.status_code}")
print(response.text)
sys.exit(1)
db_data = response.json()
print(f"Database created successfully! ID: {db_data['id']}")
return db_data['id']
def format_list_as_bullets(items):
"""Converts a python list of strings into a Notion rich_text text block with bullets."""
if not items:
return ""
text_content = ""
for item in items:
text_content += f"{item}\n"
return text_content.strip()
def get_competitor_data(data, comp_name):
"""Aggregates data from different sections of the JSON for a single competitor."""
# Init structure
comp_data = {
"name": comp_name,
"url": "",
"industries": [],
"differentiators": [],
"portfolio": [],
"silver_bullet": "",
"landmines": [],
"strengths_weaknesses": [],
"references": []
}
# 1. Basic Info & Portfolio (from 'analyses')
for analysis in data.get('analyses', []):
c = analysis.get('competitor', {})
if c.get('name') == comp_name:
comp_data['url'] = c.get('url', '')
comp_data['industries'] = analysis.get('target_industries', [])
comp_data['differentiators'] = analysis.get('differentiators', [])
# Format Portfolio
for prod in analysis.get('portfolio', []):
p_name = prod.get('product', '')
p_purpose = prod.get('purpose', '')
comp_data['portfolio'].append(f"{p_name}: {p_purpose}")
break
# 2. Battlecards
for card in data.get('battlecards', []):
if card.get('competitor_name') == comp_name:
comp_data['silver_bullet'] = card.get('silver_bullet', '')
comp_data['landmines'] = card.get('landmine_questions', [])
comp_data['strengths_weaknesses'] = card.get('strengths_vs_weaknesses', [])
break
# 3. References
for ref_entry in data.get('reference_analysis', []):
if ref_entry.get('competitor_name') == comp_name:
for ref in ref_entry.get('references', []):
r_name = ref.get('name', 'Unknown')
r_ind = ref.get('industry', '')
entry = r_name
if r_ind:
entry += f" ({r_ind})"
comp_data['references'].append(entry)
break
return comp_data
def add_competitor_entry(token, db_id, c_data):
url = "https://api.notion.com/v1/pages"
headers = {
"Authorization": f"Bearer {token}",
"Notion-Version": "2022-06-28",
"Content-Type": "application/json"
}
# Prepare properties
props = {
"Competitor Name": {"title": [{"text": {"content": c_data['name']}}]},
"USPs / Differentiators": {"rich_text": [{"text": {"content": format_list_as_bullets(c_data['differentiators'])}}]},
"Silver Bullet": {"rich_text": [{"text": {"content": c_data['silver_bullet']}}]},
"Landmines": {"rich_text": [{"text": {"content": format_list_as_bullets(c_data['landmines'])}}]},
"Strengths vs Weaknesses": {"rich_text": [{"text": {"content": format_list_as_bullets(c_data['strengths_weaknesses'])}}]},
"Portfolio": {"rich_text": [{"text": {"content": format_list_as_bullets(c_data['portfolio'])}}]},
"Known References": {"rich_text": [{"text": {"content": format_list_as_bullets(c_data['references'])}}]}
}
if c_data['url']:
props["Website"] = {"url": c_data['url']}
# Multi-select for industries
# Note: Notion options are auto-created, but we must ensure no commas or weird chars break it
ms_options = []
for ind in c_data['industries']:
# Simple cleanup
clean_ind = ind.replace(',', '')
ms_options.append({"name": clean_ind})
props["Target Industries"] = {"multi_select": ms_options}
payload = {
"parent": {"database_id": db_id},
"properties": props
}
try:
response = requests.post(url, headers=headers, json=payload)
response.raise_for_status()
print(f" - Added: {c_data['name']}")
except requests.exceptions.HTTPError as e:
print(f" - Failed to add {c_data['name']}: {e}")
# print(response.text)
def main():
token = load_notion_token(TOKEN_FILE)
data = load_json_data(JSON_FILE)
# 1. Create DB
db_id = create_competitor_database(token, PARENT_PAGE_ID)
# 2. Collect List of Competitors
# We use the shortlist or candidates list to drive the iteration
competitor_list = data.get('competitors_shortlist', [])
if not competitor_list:
competitor_list = data.get('competitor_candidates', [])
print(f"Importing {len(competitor_list)} competitors...")
for comp in competitor_list:
c_name = comp.get('name')
if not c_name: continue
# Aggregate Data
c_data = get_competitor_data(data, c_name)
# Push to Notion
add_competitor_entry(token, db_id, c_data)
print("Import complete.")
if __name__ == "__main__":
main()

230
import_relational_radar.py Normal file
View File

@@ -0,0 +1,230 @@
import json
import os
import requests
import sys
# Configuration
JSON_FILE = 'analysis_robo-planet.de.json'
TOKEN_FILE = 'notion_token.txt'
PARENT_PAGE_ID = "2e088f42-8544-8024-8289-deb383da3818"
# Database Titles
DB_TITLE_HUB = "📦 Competitive Radar (Companies)"
DB_TITLE_LANDMINES = "💣 Competitive Radar (Landmines & Intel)"
DB_TITLE_REFS = "🏆 Competitive Radar (References)"
def load_json_data(filepath):
try:
with open(filepath, 'r') as f:
return json.load(f)
except Exception as e:
print(f"Error loading JSON: {e}")
sys.exit(1)
def load_notion_token(filepath):
try:
with open(filepath, 'r') as f:
return f.read().strip()
except Exception as e:
print(f"Error loading token: {e}")
sys.exit(1)
def create_database(token, parent_page_id, title, properties):
url = "https://api.notion.com/v1/databases"
headers = {
"Authorization": f"Bearer {token}",
"Notion-Version": "2022-06-28",
"Content-Type": "application/json"
}
payload = {
"parent": {"type": "page_id", "page_id": parent_page_id},
"title": [{"type": "text", "text": {"content": title}}],
"properties": properties
}
response = requests.post(url, headers=headers, json=payload)
if response.status_code != 200:
print(f"Error creating DB '{title}': {response.status_code}")
print(response.text)
sys.exit(1)
db_data = response.json()
print(f"✅ Created DB '{title}' (ID: {db_data['id']})")
return db_data['id']
def create_page(token, db_id, properties):
url = "https://api.notion.com/v1/pages"
headers = {
"Authorization": f"Bearer {token}",
"Notion-Version": "2022-06-28",
"Content-Type": "application/json"
}
payload = {
"parent": {"database_id": db_id},
"properties": properties
}
response = requests.post(url, headers=headers, json=payload)
if response.status_code != 200:
print(f"Error creating page: {response.status_code}")
# print(response.text)
return None
return response.json()['id']
def format_list_as_bullets(items):
if not items: return ""
return "\n".join([f"{item}" for item in items])
def main():
token = load_notion_token(TOKEN_FILE)
data = load_json_data(JSON_FILE)
print("🚀 Starting Relational Import...")
# --- STEP 1: Define & Create Competitors Hub DB ---
props_hub = {
"Name": {"title": {}},
"Website": {"url": {}},
"Target Industries": {"multi_select": {}},
"Portfolio Summary": {"rich_text": {}},
"Silver Bullet": {"rich_text": {}},
"USPs": {"rich_text": {}}
}
hub_db_id = create_database(token, PARENT_PAGE_ID, DB_TITLE_HUB, props_hub)
# --- STEP 2: Define & Create Satellite DBs (Linked to Hub) ---
# Landmines DB
props_landmines = {
"Statement / Question": {"title": {}},
"Type": {"select": {
"options": [
{"name": "Landmine Question", "color": "red"},
{"name": "Competitor Weakness", "color": "green"},
{"name": "Competitor Strength", "color": "orange"}
]
}},
"Related Competitor": {
"relation": {
"database_id": hub_db_id,
"dual_property": {"synced_property_name": "Related Landmines & Intel"}
}
}
}
landmines_db_id = create_database(token, PARENT_PAGE_ID, DB_TITLE_LANDMINES, props_landmines)
# References DB
props_refs = {
"Customer Name": {"title": {}},
"Industry": {"select": {}},
"Snippet": {"rich_text": {}},
"Case Study URL": {"url": {}},
"Related Competitor": {
"relation": {
"database_id": hub_db_id,
"dual_property": {"synced_property_name": "Related References"}
}
}
}
refs_db_id = create_database(token, PARENT_PAGE_ID, DB_TITLE_REFS, props_refs)
# --- STEP 3: Import Competitors (and store IDs) ---
competitor_map = {} # Maps Name -> Notion Page ID
competitors = data.get('competitors_shortlist', []) or data.get('competitor_candidates', [])
print(f"\nImporting {len(competitors)} Competitors...")
for comp in competitors:
c_name = comp.get('name')
if not c_name: continue
# Gather Data
c_url = comp.get('url', '')
# Find extended analysis data
analysis_data = next((a for a in data.get('analyses', []) if a.get('competitor', {}).get('name') == c_name), {})
battlecard_data = next((b for b in data.get('battlecards', []) if b.get('competitor_name') == c_name), {})
industries = analysis_data.get('target_industries', [])
portfolio = analysis_data.get('portfolio', [])
portfolio_text = "\n".join([f"{p.get('product')}: {p.get('purpose')}" for p in portfolio])
usps = format_list_as_bullets(analysis_data.get('differentiators', []))
silver_bullet = battlecard_data.get('silver_bullet', '')
# Create Page
props = {
"Name": {"title": [{"text": {"content": c_name}}]},
"Portfolio Summary": {"rich_text": [{"text": {"content": portfolio_text[:2000]}}]},
"USPs": {"rich_text": [{"text": {"content": usps[:2000]}}]},
"Silver Bullet": {"rich_text": [{"text": {"content": silver_bullet[:2000]}}]},
"Target Industries": {"multi_select": [{"name": i.replace(',', '')} for i in industries]},
}
if c_url: props["Website"] = {"url": c_url}
page_id = create_page(token, hub_db_id, props)
if page_id:
competitor_map[c_name] = page_id
print(f" - Created: {c_name}")
# --- STEP 4: Import Landmines & Intel ---
print("\nImporting Landmines & Intel...")
for card in data.get('battlecards', []):
c_name = card.get('competitor_name')
comp_page_id = competitor_map.get(c_name)
if not comp_page_id: continue
# 1. Landmines
for q in card.get('landmine_questions', []):
props = {
"Statement / Question": {"title": [{"text": {"content": q}}]},
"Type": {"select": {"name": "Landmine Question"}},
"Related Competitor": {"relation": [{"id": comp_page_id}]}
}
create_page(token, landmines_db_id, props)
# 2. Weaknesses
# The JSON has "strengths_vs_weaknesses" combined. We'll import them as general Intel points.
for point in card.get('strengths_vs_weaknesses', []):
# Try to guess type based on text, or just default to Weakness context from Battlecard
p_type = "Competitor Weakness" # Assuming these are points for us to exploit
props = {
"Statement / Question": {"title": [{"text": {"content": point}}]},
"Type": {"select": {"name": p_type}},
"Related Competitor": {"relation": [{"id": comp_page_id}]}
}
create_page(token, landmines_db_id, props)
print(" - Landmines imported.")
# --- STEP 5: Import References ---
print("\nImporting References...")
count_refs = 0
for ref_group in data.get('reference_analysis', []):
c_name = ref_group.get('competitor_name')
comp_page_id = competitor_map.get(c_name)
if not comp_page_id: continue
for ref in ref_group.get('references', []):
r_name = ref.get('name', 'Unknown')
r_industry = ref.get('industry', 'Unknown')
r_snippet = ref.get('testimonial_snippet', '')
r_url = ref.get('case_study_url', '')
props = {
"Customer Name": {"title": [{"text": {"content": r_name}}]},
"Industry": {"select": {"name": r_industry}},
"Snippet": {"rich_text": [{"text": {"content": r_snippet[:2000]}}]},
"Related Competitor": {"relation": [{"id": comp_page_id}]}
}
if r_url and r_url.startswith('http'):
props["Case Study URL"] = {"url": r_url}
create_page(token, refs_db_id, props)
count_refs += 1
print(f" - {count_refs} References imported.")
print("\n✅ Relational Import Complete!")
if __name__ == "__main__":
main()

35
refresh_references.py Normal file
View File

@@ -0,0 +1,35 @@
import asyncio
import json
import os
import sys
# Path to the orchestrator
sys.path.append(os.path.join(os.getcwd(), 'competitor-analysis-app'))
from competitor_analysis_orchestrator import analyze_single_competitor_references
async def refresh_references():
json_path = 'analysis_robo-planet.de.json'
with open(json_path, 'r') as f:
data = json.load(f)
competitors = data.get('competitors_shortlist', [])
if not competitors:
competitors = data.get('competitor_candidates', [])
print(f"Refreshing references for {len(competitors)} competitors...")
tasks = [analyze_single_competitor_references(c) for c in competitors]
results = await asyncio.gather(*tasks)
# Filter and update
data['reference_analysis'] = [r for r in results if r is not None]
with open(json_path, 'w') as f:
json.dump(data, f, indent=2)
print(f"Successfully updated {json_path} with grounded reference data.")
if __name__ == "__main__":
asyncio.run(refresh_references())