HOALedgerIQ_Website/agents/marketing-seo/scripts/ga4-direct.py

#!/usr/bin/env python3
"""
Google Analytics 4 - Direct JWT Authentication (No gcloud required)
Enhanced to pull: sessions, users, engagement, scroll depth, form interactions
"""
import json
from datetime import datetime, timedelta
from pathlib import Path
from google.analytics.data import BetaAnalyticsDataClient
from google.analytics.data_v1beta.types import RunReportRequest, DateRange, Metric, Dimension, Filter, FilterExpression

CONFIG_DIR = Path(__file__).parent.parent / "config"
GA_CREDENTIALS = CONFIG_DIR / "ga-credentials.json"
GA_PROPERTY_ID = "526394825"

def load_credentials():
    """Load service account credentials"""
    with open(GA_CREDENTIALS) as f:
        return json.load(f)

def get_engagement_data():
    """Get comprehensive engagement metrics"""
    try:
        client = BetaAnalyticsDataClient.from_service_account_json(str(GA_CREDENTIALS))
        
        # Main traffic report
        request = RunReportRequest(
            property=f"properties/{GA_PROPERTY_ID}",
            date_ranges=[DateRange(start_date="1daysAgo", end_date="today")],
            metrics=[
                Metric(name="sessions"),
                Metric(name="activeUsers"),
                Metric(name="newUsers"),
                Metric(name="bounceRate"),
                Metric(name="averageSessionDuration"),
                Metric(name="engagementRate"),
                Metric(name="screenPageViews"),
            ]
        )
        
        response = client.run_report(request)
        
        result = {
            "success": True,
            "timestamp": datetime.now().isoformat()
        }
        
        for row in response.rows:
            result['sessions'] = int(row.metric_values[0].value)
            result['activeUsers'] = int(row.metric_values[1].value)
            result['newUsers'] = int(row.metric_values[2].value)
            result['bounceRate'] = float(row.metric_values[3].value)
            result['averageSessionDuration'] = float(row.metric_values[4].value)
            result['engagementRate'] = float(row.metric_values[5].value)
            result['screenPageViews'] = int(row.metric_values[6].value)
        
        # Get scroll depth events
        scroll_request = RunReportRequest(
            property=f"properties/{GA_PROPERTY_ID}",
            dimensions=[Dimension(name="eventName")],
            metrics=[Metric(name="eventCount")],
            date_ranges=[DateRange(start_date="1daysAgo", end_date="today")],
            dimension_filter=FilterExpression(
                filter=Filter(
                    field_name="eventName",
                    string_filter=Filter.StringFilter(
                        match_type=Filter.StringFilter.MatchType.BEGINS_WITH,
                        value="scroll"
                    )
                )
            ),
            limit=10
        )
        
        scroll_response = client.run_report(scroll_request)
        scroll_data = {}
        for row in scroll_response.rows:
            event_name = row.dimension_values[0].value
            count = int(row.metric_values[0].value)
            scroll_data[event_name] = count
        
        result['scroll_25'] = scroll_data.get('scroll_25', 0)
        result['scroll_50'] = scroll_data.get('scroll_50', 0)
        result['scroll_75'] = scroll_data.get('scroll_75', 0)
        result['scroll_total'] = sum(scroll_data.values())
        
        # Get form interactions
        form_request = RunReportRequest(
            property=f"properties/{GA_PROPERTY_ID}",
            metrics=[Metric(name="eventCount")],
            date_ranges=[DateRange(start_date="1daysAgo", end_date="today")],
            dimension_filter=FilterExpression(
                filter=Filter(
                    field_name="eventName",
                    string_filter=Filter.StringFilter(
                        match_type=Filter.StringFilter.MatchType.EXACT,
                        value="form_start"
                    )
                )
            )
        )
        
        form_response = client.run_report(form_request)
        form_count = 0
        for row in form_response.rows:
            form_count += int(row.metric_values[0].value)
        
        result['form_interactions'] = form_count
        
        # Get click events (outbound clicks)
        click_request = RunReportRequest(
            property=f"properties/{GA_PROPERTY_ID}",
            metrics=[Metric(name="eventCount")],
            date_ranges=[DateRange(start_date="1daysAgo", end_date="today")],
            dimension_filter=FilterExpression(
                filter=Filter(
                    field_name="eventName",
                    string_filter=Filter.StringFilter(
                        match_type=Filter.StringFilter.MatchType.EXACT,
                        value="click"
                    )
                )
            )
        )
        
        click_response = client.run_report(click_request)
        click_count = 0
        for row in click_response.rows:
            click_count += int(row.metric_values[0].value)
        
        result['outbound_clicks'] = click_count
        
        return result
        
    except Exception as e:
        return {"success": False, "error": str(e)}

if __name__ == "__main__":
    print("🚀 Fetching GA4 Analytics Data...\n")
    
    result = get_engagement_data()
    
    if result.get('success'):
        print("📊 Traffic (Last 24h):")
        print(f"  ✅ Sessions: {result.get('sessions', 0):,}")
        print(f"  ✅ Active Users: {result.get('activeUsers', 0):,}")
        print(f"  ✅ New Users: {result.get('newUsers', 0):,}")
        print(f"  ✅ Engagement Rate: {result.get('engagementRate', 0):.1%}")
        print(f"  ✅ Bounce Rate: {result.get('bounceRate', 0):.1%}")
        print(f"  ✅ Avg Session: {result.get('averageSessionDuration', 0):.1f}s")
        print(f"  ✅ Page Views: {result.get('screenPageViews', 0):,}")
        
        print("\n🎯 Engagement Events:")
        print(f"  • Scroll Depth 25%: {result.get('scroll_25', 0)}")
        print(f"  • Scroll Depth 50%: {result.get('scroll_50', 0)}")
        print(f"  • Scroll Depth 75%: {result.get('scroll_75', 0)}")
        print(f"  • Form Interactions: {result.get('form_interactions', 0)}")
        print(f"  • Outbound Clicks: {result.get('outbound_clicks', 0)}")
        
    else:
        print(f"❌ Error: {result.get('error')}")