Analytics Funnel: Track Widget Lifecycle (Loads → Opens → Conversations → Conversions)

Problem

Currently, we have no visibility into how visitors interact with the chat widget before they start a conversation. We don't know:
- How many people see the widget on the page
- - How many click to open the chat
- - What percentage actually start a conversation
- - Where the biggest drop-offs occur

This makes it impossible to optimize the widget placement, design, or initial prompts to improve engagement.

Proposed Solution

Implement a complete widget lifecycle funnel that tracks four key stages:

1. Widget Loads - When the widget script loads on a page
2. Chat Opened - When a user clicks the toggle to open the chat
3. Conversation Started - When the user sends their first message
4. Conversion - When a user completes an action (contact form, newsletter, etc.)

Features

• New WidgetEvent model to track load and open events with session deduplication
• • API endpoints for tracking widget lifecycle events
• • Dashboard funnel card with animated visualization showing drop-off rates
• • Analytics page detailed funnel with trend charts over time
• • Daily/hourly aggregation for historical analysis
• • Automatic cleanup of raw events after 30 days (aggregated data preserved)

Benefits

• Identify where users drop off in the funnel
• • A/B test widget designs and placements
• • Track improvement over time
• • Better understand user engagement patterns