Google Analytics (GA4)

7 production-ready jobs — demographic personas, interest-based content strategy, funnel focus groups, landing page refresh, real-time spike response, channel-persona mapping, and device-aware creative recommendations

Google Analytics 4 holds your audience truth — who visits, where they come from, what they do, and where they drop off. These seven jobs pull that data through Mavera’s surfaces (Custom Personas, Mave Agent, Focus Groups, Generate) so your personas, content, and creative are grounded in real behavioral data instead of assumptions.

---

API Reference Card

Detail	Value
Base URL	https://analyticsdata.googleapis.com/v1beta
Auth	OAuth 2.0 via service account (GOOGLE_APPLICATION_CREDENTIALS JSON key file)
Rate limits	50,000 requests/project/day; 10 concurrent requests/property
Python client	google-analytics-data — official GA4 Data API client
Mavera base	https://app.mavera.io/api/v1
Mavera auth	Authorization: Bearer mvra_live_...

All examples use three environment variables: GOOGLE_APPLICATION_CREDENTIALS (path to service account JSON), GA4_PROPERTY_ID (numeric property ID, e.g. 123456789), and MAVERA_API_KEY. Python examples use the google-analytics-data client library; JavaScript examples use the REST API directly with a service account access token.

---

Prerequisites

1. GA4 property with data

   You need an active GA4 property with at least 30 days of traffic. Find the property ID in GA4 → Admin → Property Settings.
2. Service account with GA4 access

   Create a service account in Google Cloud, download the JSON key, then add the service account email as a Viewer in GA4 → Admin → Property Access Management.
3. Enable the Analytics Data API

   In Google Cloud Console, navigate to APIs & Services → Enable the Google Analytics Data API.
4. Install dependencies

   pip install google-analytics-data requests
   
5. Environment variables

   export GOOGLE_APPLICATION_CREDENTIALS="/path/to/service-account.json"
   export GA4_PROPERTY_ID="123456789"
   export MAVERA_API_KEY="mvra_live_xxxxx"
   

---

Jobs

#	Job	GA4 Data	Mavera Surface	Output
1	Audience Demographics → Persona Creation	RunReport: age, gender, country, device, conversions	Custom Personas	High-value intersection personas
2	Interest Category → Content Strategy	Interests dimension report	Mave Agent	Content plan bridging product + interests
3	Conversion Path → Focus Group Validation	Funnel report with step drop-off	Focus Groups	Persona feedback on drop-off points
4	Landing Page Performance → Content Refresh	Top landing pages with bounce + engagement	Generate	Audience-specific copy rewrites
5	Real-Time Audience → Trending Response	Real-time active users by page/source	Mave Agent	Spike diagnosis + response plan
6	Acquisition Channel × Persona Mapping	Session source/medium × demographics	Custom Personas	Channel-demographic persona map
7	Device Behavior → Creative Format Recommendations	Device category, resolution, engagement	Mave Agent	Format adjustments per device

---

Rate Limits & Production Notes

GA4 API Endpoint	Limit	Strategy
RunReport	10 concurrent per property; 50,000/day	Queue reports; use date range caching
RunRealtimeReport	Same quota pool	Poll at max 1/minute for dashboards
RunFunnelReport	Same quota pool	Cache results; funnels rarely need real-time refresh

The GA4 Data API enforces 10 concurrent requests per property. If you run jobs 1–7 in parallel, you’ll exceed this limit. Run sequentially or batch with 300ms delays. Retry on 429 with exponential backoff.

Production checklist:

Concern	Recommendation
Service account key rotation	Rotate keys every 90 days via Google Cloud IAM. Store in a secrets manager, never in code.
Date ranges	Always use relative ranges (30daysAgo) for scheduled jobs. Absolute dates risk stale data.
Data freshness	Standard GA4 properties have 24–48h processing delay for non-real-time reports. Real-time reports cover the last 30 minutes only.
Sampling	Reports over 10M events may be sampled. Use shorter date ranges or the GA4 BigQuery export for unsampled data.
Mavera credits	Monitor at Dashboard. Mave chat calls consume credits proportional to prompt length.
Thresholding	GA4 redacts rows with small user counts to protect privacy. If critical segments are missing, reduce dimension cardinality or extend date ranges.

---

What’s Next

All Integrations

50+ API integrations with full code

GA4 Data API Docs

Official GA4 Data API reference

Personas API

Full reference for POST /api/v1/personas

Focus Groups API

Full reference for POST /api/v1/focus-groups

Mave Agent

Full reference for POST /api/v1/mave/chat

Generate API

Full reference for POST /api/v1/generations