How to Install and Use Google Analytics in OpenClaw

Open your terminal and run the ClawHub install command. The google-analytics skill provides natural language querying of Google Analytics 4 data via the official GA4 Data API. After installation, ClawHub shows the required configuration: GOOGLE_ANALYTICS_PROPERTY_ID (your GA4 property ID) and a Google service account credentials file path. If you are familiar with other OpenClaw skills, note that this skill uses a service account JSON file for authentication rather than a simple API key string. This is a Google requirement for server-to-server API access — your service account authenticates with Google using a public/private key pair contained in the JSON file. The setup takes a few extra minutes but only needs to be done once. Run `clawhub info google-analytics` after installation to see all available configuration options, including support for multiple GA4 properties and custom dimension mappings.

1clawhub install google-analytics

Your GOOGLE_ANALYTICS_PROPERTY_ID is the unique identifier for your GA4 property. It is a numeric ID, not the tracking ID starting with 'G-' (that is the Measurement ID for client-side tracking, not the property ID for API access). To find your property ID: Open Google Analytics > click Admin (gear icon, bottom-left) > under the Property column, click Property Settings. The Property ID is shown at the top of the settings page as a 9-10 digit number like `123456789`. Write this number down — you will set it as GOOGLE_ANALYTICS_PROPERTY_ID. When making API requests, the property is referenced as `properties/123456789`, but the skill handles this format automatically — you just provide the numeric ID.

Google requires server-to-server API access to use service accounts rather than user OAuth tokens. A service account is a non-human Google account that your application uses to access APIs on your behalf. Go to console.cloud.google.com and select (or create) a Google Cloud project. Navigate to APIs & Services > Library and search for 'Google Analytics Data API v1'. Click it and enable it for your project. Next, go to IAM & Admin > Service Accounts. Click 'Create Service Account'. Give it a name like 'openclaw-analytics'. Click through the role assignment screen without adding a role (the permission will be granted in GA4 directly in the next step). Click Done. Click the new service account in the list to open it. Go to the Keys tab > Add Key > Create New Key > JSON. Download the JSON key file. This file contains the private key the skill uses for authentication. Save it to `~/.openclaw/credentials/google-analytics.json` — create the `credentials` directory if it does not exist. Note the service account email address (format: `openclaw-analytics@your-project.iam.gserviceaccount.com`) — you need it for the next step when granting GA4 access.

1# Create the credentials directory and move the downloaded key file
2mkdir -p ~/.openclaw/credentials
3mv ~/Downloads/your-project-XXXXXXXX.json ~/.openclaw/credentials/google-analytics.json
4
5# Verify the file is in place
6ls -la ~/.openclaw/credentials/google-analytics.json

The service account needs to be granted access to your GA4 property before it can read data. This is done in the Google Analytics interface, not in Google Cloud. Open Google Analytics > Admin > Property Access Management (under the Property column). Click the + button to add a user. Enter the service account email address (the `...@...iam.gserviceaccount.com` address from Step 3). Set the role to 'Viewer' — this grants read-only access to GA4 data, which is the minimum required for the skill. Click 'Add'. Now configure OpenClaw with your property ID and credentials file path. Run the config commands below, then reload OpenClaw. The skill will test the connection on first use by making a simple metadata request to the GA4 API.

1# Set your GA4 Property ID
2clawhub config set GOOGLE_ANALYTICS_PROPERTY_ID 123456789
3
4# Set the path to your service account credentials file
5clawhub config set GOOGLE_APPLICATION_CREDENTIALS ~/.openclaw/credentials/google-analytics.json
6
7# Apply changes
8clawhub reload
9
10# Verify both are set
11clawhub config get GOOGLE_ANALYTICS_PROPERTY_ID
12clawhub config get GOOGLE_APPLICATION_CREDENTIALS

With authentication configured, open OpenClaw chat and ask your first analytics question. The skill translates natural language queries into GA4 Data API requests, selecting the correct metrics, dimensions, date ranges, and filters based on your question. The skill understands relative date language ('last week', 'yesterday', 'last 30 days', 'this month vs last month') and translates it to the correct GA4 date range format. It handles comparison queries ('how does this week compare to last week'), filtering ('only show mobile users', 'only from organic search'), and sorting ('sorted by sessions descending'). For complex analyses, be specific about what you want to see. 'Show me my traffic' is too vague — 'show me daily sessions for the last 2 weeks, broken down by device category' gives the skill enough context to build the precise API request. The more specific your query, the more useful the output. RapidDev recommends setting up a morning briefing prompt as a saved OpenClaw command — a quick daily analytics summary that runs automatically and delivers key metrics without any manual effort.

1# Optional: ~/.openclaw/skills/google-analytics.yaml for advanced config
2google-analytics:
3  property_id: ""         # Leave blank if using GOOGLE_ANALYTICS_PROPERTY_ID env var
4  credentials_file: ""    # Leave blank if using GOOGLE_APPLICATION_CREDENTIALS env var
5  default_date_range: "last30Days"   # Default when no date is specified in prompts
6  max_rows: 100           # Maximum rows returned per query (default: 50)
7  # Cache responses to reduce API quota usage
8  cache_ttl: 300          # Seconds to cache identical queries (default: 5 min, 0 = no cache)