🔍 Churn Scout - Market Intelligence Agent

Autonomous AI agent that reveals why customers are leaving your competitors

---

🎯 What is Churn Scout?

Churn Scout is an autonomous market intelligence agent designed for SaaS founders and marketing teams. It automatically identifies why customers are leaving a competitor by analyzing public sentiment signals.

Unlike traditional scrapers that simply dump raw text, Churn Scout uses an internal Machine Learning Engine to cluster thousands of complaints into specific Pain Points (e.g., "Pricing is too high," "Mobile app crashes").

✨ Key Features

Feature	Description
🤖 Zero API Keys	No OpenAI, no Gemini, no external AI costs
🧠 ML-Powered Analysis	Scikit-Learn clusters complaints into actionable topics
📊 Beautiful Dashboard	Self-contained HTML report with interactive insights
🔒 Privacy-First	Aggregates data, never stores PII
⚡ Fast Insights	Get competitor intelligence in minutes

---

🚀 How It Works

graph LR
    A[Input: Competitor Name] --> B[Playwright Scraper]
    B --> C[Reddit Search]
    C --> D[TextBlob Sentiment]
    D --> E[TF-IDF + K-Means]
    E --> F[Pain Point Clusters]
    F --> G[Interactive Dashboard]

1. Visual Scraping: Playwright navigates public Reddit search to find complaints
2. Sentiment Filtering: TextBlob filters for negative sentiment (churn signals)
3. AI Clustering: Scikit-Learn groups similar complaints into topics
4. Smart Reporting: Generates a hosted HTML dashboard with actionable insights

---

📥 Input Configuration

Field	Type	Description	Default
competitorName	String	The brand to analyze (e.g., "Notion", "Jira")	Required
maxPosts	Integer	Sample size (50-500). Higher = more accurate	100
proxyConfiguration	Object	Apify Proxy settings	Enabled

Example Input

{
    "competitorName": "Slack",
    "maxPosts": 200,
    "proxyConfiguration": { "useApifyProxy": true }
}

---

📤 Output

1. Interactive Dashboard (HTML)

A beautiful, self-contained dashboard stored in the Key-Value Store:

• 📊 Churn Signal Count - Total negative mentions found
• 📉 Average Sentiment - Overall negativity score
• 🏷️ Pain Point Clusters - AI-identified complaint categories
• 📝 Raw Evidence - Original posts with source links

2. Structured Dataset (JSON)

[
    {
        "text": "Slack's pricing is ridiculous for small teams",
        "topic": "ISSUE: PRICING, EXPENSIVE, TEAMS",
        "polarity": -0.42,
        "url": "https://reddit.com/r/..."
    }
]

---

🛠️ Technology Stack

Layer	Technology	Purpose
Scraping	Playwright	Visual browser automation
NLP	TextBlob	Sentiment polarity analysis
ML	Scikit-Learn	TF-IDF + K-Means clustering
Templating	Jinja2	Dashboard generation
Infrastructure	Docker + Apify	Serverless execution

---

💼 Use Cases

• 🎯 Competitive Intelligence - Understand competitor weaknesses
• 📢 Marketing Strategy - Craft messaging that addresses pain points
• 🛠️ Product Development - Build features competitors lack
• 📈 Sales Enablement - Arm sales team with competitor objections

---

📋 Compliance & Ethics

✅ Rate Limited - Mimics human browsing speed
✅ Public Data Only - Only accesses reddit.com/search
✅ No PII - Aggregates into clusters, no individual targeting
✅ Transformative - Produces insights, not raw data dumps

---

💰 Pricing

Model	Price
Recurring Rental	$20/month
Pay-Per-Run	$0.50/run

---

🤝 Support

• 📧 Issues? Open a ticket on Apify
• ⭐ Love it? Leave a review!

---