Analytics Stats#

Endpoints for querying aggregated analytics data. Provides user engagement funnels, popup performance, purchase/revenue metrics with A/B testing, overview stats, and AI-powered conversation insights.

Common request parameters#

Used by overview, engagement, popup-metrics, and purchases endpoints:

All time-based queries use: start_date = now() - timeframe days, end_date = now().

Endpoints#

GET /api/dashboard/analytics/stats/overview#

High-level conversation statistics.

Response:

{
  "timeframe_days": 7,
  "start_date": "2026-02-17",
  "end_date": "2026-02-24",
  "total_conversations": 150,
  "total_ai_messages": 500,
  "saved_hours": "25.5"
}

GET /api/dashboard/analytics/stats/engagement#

Visitor funnel: visitors -> approached (popup shown) -> opened (widget opened) -> chatted (conversation started).

Response:

{
  "timeframe_days": 7,
  "start_date": "2026-02-17",
  "end_date": "2026-02-24",
  "total_users": 1250,
  "total_users_approached": 1050,
  "total_users_opened": 950,
  "total_users_chatted": 800,
  "approached_ratio": 84.0,
  "opened_ratio": 76.0,
  "chatted_ratio": 64.0
}

GET /api/dashboard/analytics/stats/popup-metrics#

Popup performance grouped by version, plus conversation source breakdown.

Response:

{
  "timeframe_days": 7,
  "start_date": "2026-02-17",
  "end_date": "2026-02-24",
  "metrics": [
    {
      "popup_version": "v1",
      "popup_displayed_count": 1000,
      "popup_opened_count": 250,
      "open_rate": 25.0
    },
    {
      "popup_version": "v2",
      "popup_displayed_count": 800,
      "popup_opened_count": 300,
      "open_rate": 37.5
    }
  ],
  "conversations": {
    "total": 450,
    "from_popup": 300,
    "from_chat_button": 150,
    "popup_ratio": 66.67
  }
}

Popup versions sorted alphabetically. Unknown versions grouped as "unknown".

GET /api/dashboard/analytics/stats/purchases#

Purchase and revenue metrics, with chat vs. no-chat comparison. Optionally includes A/B test breakdown.

Response (without A/B test):

{
  "timeframe_days": 7,
  "start_date": "2026-02-17",
  "end_date": "2026-02-24",
  "total_users": 500,
  "total_users_chatted": 200,
  "total_users_purchased": 100,
  "total_users_chatted_purchased": 80,
  "total_revenue_purchased": 25000.00,
  "total_revenue_chatted_purchased": 20000.00,
  "conversion_rate_with_chat": 40.0,
  "conversion_rate_without_chat": 4.0,
  "aov_with_chat": 250.0,
  "aov_without_chat": 250.0
}

"Purchase after chat" logic:

1. Find earliest conversation start event per user
2. Count purchase events occurring after that time
3. Only purchases after the user's first chat count as "chatted_purchased"

Purchase deduplication: Duplicate order numbers are skipped. Only the first event per order is counted.

Revenue: Uses EUR-converted values for cross-currency aggregation.

Response with A/B test (when A/B testing is enabled for the chatbot):

{
  "...": "same fields as above",
  "ab_test": {
    "control": {
      "visitors": 250,
      "orders": 50,
      "conversion_rate": 8.0,
      "aov": 250.0,
      "arpu": 50.0
    },
    "test": {
      "visitors": 250,
      "orders": 60,
      "conversion_rate": 10.0,
      "aov": 260.0,
      "arpu": 62.4
    }
  }
}

Group assignment is derived from the user token's last character (see Event Tracking).

GET /api/dashboard/analytics/stats/conversation-insights#

AI-powered conversation analysis report. Aggregates structured analysis data from analyzed conversations.

Request parameters (different from other endpoints):

Timeframe formats:

Response:

{
  "most_clicked_products": [
    {
      "item": "https://example.com/product/123",
      "count": 45,
      "conversations": [
        "https://app.selzee.com/app/{chatbot_id}/conversations/{conv_id}"
      ]
    }
  ],
  "most_talked_about_products": [
    { "item": "Nike Air Max 90", "count": 30, "conversations": ["..."] }
  ],
  "most_frequent_intents": [
    { "item": "Product availability question", "count": 25, "conversations": ["..."] }
  ],
  "most_common_complaints": [
    { "item": "Shipping delays", "count": 8, "conversations": ["..."] }
  ],
  "mood_percentages": {
    "happy": 65.5,
    "angry": 10.2,
    "neutral": 24.3
  },
  "outcome_percentages": {
    "success": 75.0,
    "fail": 10.0,
    "neutral": 15.0
  },
  "human_intervention_needed_percentage": 12.5,
  "missing_information_overview": [
    { "item": "Product availability in size L", "count": 8, "conversations": ["..."] }
  ],
  "total_conversations_analyzed": 100,
  "total_click_throughs": 145,
  "chat_conversion": {
    "ratio": 35.0,
    "total_conversions": 35,
    "total_conversations": 100
  },
  "reactions": {
    "likes": 200,
    "dislikes": 15
  },
  "average_conversation_duration_minutes": 3.5
}

Ranked item format: { item: string, count: int, conversations: string[] } — sorted by count descending, with links to each relevant conversation.

Limits: Max 100 conversations per report.

Error handling#