Stop checking dashboards. Let your AI assistant tell you what's happening with your website.

By Chief Wizard & James Tipton   |  Category: AI Engineering

If you're running a business, blog, or brand online, you know the drill: log into Google Analytics, squint at charts, try to remember what last week's numbers were, and wonder if anyone's actually reading your content.

What if your AI assistant just told you?

In this guide we'll connect Google Analytics to your OpenClaw bot so it can report your daily views, sessions, and traffic sources on command — alert you to spikes when something goes viral — deliver scheduled reports right to your Telegram — and track multiple websites from one place. No more logging into dashboards. Your bot becomes your personal data analyst.

1 Set Up a Google Cloud Project

First we need to create a project in Google Cloud that gives your bot permission to read analytics data.

1. Go to console.cloud.google.com
2. Click Select a Project → New Project
3. Name it something like MyBot-Analytics
4. Click Create

Enable the Analytics API

1. In the search bar type: Google Analytics Data API
2. Click it when it appears
3. Click the blue Enable button

2 Create a Service Account

A service account is like a robot user that your bot authenticates as when it talks to Google. Think of it as creating a read-only employee that only your bot knows the password to.

1. Go to the hamburger menu (☰) → IAM & Admin → Service Accounts
2. Click + Create Service Account
3. Name it: analytics-reader
4. Click Create and Continue
5. For role, select Viewer or skip
6. Click Done

Download the credentials key

1. Click on the service account you just created
2. Go to the Keys tab
3. Click Add Key → Create new key
4. Choose JSON
5. Click Create — a JSON file will download automatically

3 Set Up Google Analytics

If you don't have Google Analytics on your site yet:

1. Go to analytics.google.com
2. Click Start Measuring
3. Create an account and property for your website
4. Complete the setup flow
5. When you get the tracking code (starts with G-), add it to your website's header

4 Grant Your Bot Access to Analytics

Now connect the service account to your Google Analytics property so it can actually read your data.

1. Go to analytics.google.com
2. Click Admin (gear icon, bottom left)
3. Click Property Access Management
4. Click + → Add users
5. Paste your service account email — it looks like: [email protected]
6. Set role to Viewer
7. Click Add

Get your Property ID

While you're in Admin:

1. Click Property Settings
2. Find your Property ID — a number like 123456789
3. Save this — your bot needs it in the next step

5 Store Credentials on Your Server

SSH into your OpenClaw server and create a secure home for the credentials file:

mkdir -p ~/analytics/credentials
chmod 700 ~/analytics/credentials

Copy your downloaded JSON key to the server. From your Mac:

scp ~/Downloads/your-key-file.json parallels@YOUR_VM_IP:~/analytics/credentials/ga-key.json

Lock down the permissions:

chmod 600 ~/analytics/credentials/ga-key.json

6 Create the Analytics Script

Install the Google Analytics Python library:

pip install google-analytics-data --break-system-packages

Create the analytics script:

nano ~/analytics/ga_report.py

Paste this:

import os
from google.analytics.data_v1beta import BetaAnalyticsDataClient
from google.analytics.data_v1beta.types import (
    DateRange, Dimension, Metric, RunReportRequest
)
 
KEY_PATH = os.path.expanduser("~/analytics/credentials/ga-key.json")
os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = KEY_PATH
 
# Add your property IDs here
PROPERTIES = {
    "main-site": "YOUR_PROPERTY_ID",
}
 
def get_report(property_name="main-site", days=7):
    property_id = PROPERTIES.get(property_name)
    if not property_id:
        return f"Unknown property: {property_name}"
 
    client = BetaAnalyticsDataClient()
    request = RunReportRequest(
        property=f"properties/{property_id}",
        date_ranges=[DateRange(start_date=f"{days}daysAgo", end_date="today")],
        dimensions=[Dimension(name="date")],
        metrics=[
            Metric(name="sessions"),
            Metric(name="activeUsers"),
            Metric(name="screenPageViews"),
        ],
    )
    response = client.run_report(request)
 
    total_sessions = 0
    total_users = 0
    total_views = 0
 
    for row in response.rows:
        total_sessions += int(row.metric_values[0].value)
        total_users += int(row.metric_values[1].value)
        total_views += int(row.metric_values[2].value)
 
    return (
        f"📊 Analytics Report — Last {days} days ({property_name})\n"
        f"Sessions: {total_sessions:,}\n"
        f"Users: {total_users:,}\n"
        f"Page Views: {total_views:,}"
    )
 
if __name__ == "__main__":
    print(get_report())

Replace YOUR_PROPERTY_ID with the number you saved in Step 4. Test it:

python3 ~/analytics/ga_report.py

You should see a formatted report with your last 7 days of data.

7 Create the OpenClaw Skill

Now teach your bot how to use the script. Create a skill file:

mkdir -p ~/.openclaw/workspace/skills/google-analytics
nano ~/.openclaw/workspace/skills/google-analytics/SKILL.md

Paste this:

---
name: google-analytics
description: Fetch Google Analytics traffic data for tracked websites
version: 1.0.0
---
 
# Google Analytics Skill
 
You have access to a Google Analytics reporting script.
 
## How to run a report
 
Use your bash tool to run:
 
python3 ~/analytics/ga_report.py
 
This returns sessions, users, and page views for the last 7 days.
 
## When to use this skill
 
- When asked about traffic, views, sessions, or analytics
- When asked "how is the site doing?"
- During scheduled analytics report cron jobs
- When a traffic spike alert is triggered
 
## Interpreting results
 
- Sessions = visits to the site
- Users = unique visitors
- Page Views = total pages loaded
 
Compare to previous periods to identify trends.
If numbers look unusual, mention it and suggest investigating traffic sources.

Save with Ctrl+O → Enter → Ctrl+X. Your bot will pick up the skill on the next session.

8 Set Up Automated Reports with Cron

Now make your bot deliver reports automatically. Use OpenClaw's cron system:

openclaw cron add \
  --name "Daily Analytics Report" \
  --cron "0 9 * * *" \
  --message "Run the analytics report and share the results" \
  --channel telegram \
  --to YOUR_TELEGRAM_CHAT_ID

This sends a report every morning at 9 AM. Change the 9 to your preferred hour in 24-hour format. Use your timezone setting if your OpenClaw supports it — otherwise convert to UTC.

9 Add Heartbeat Monitoring

Want your bot to proactively alert you when something interesting happens — without you having to ask? Add analytics checking to your heartbeat file.

Edit ~/.openclaw/workspace/HEARTBEAT.md and add:

# Analytics Monitoring
 
## Periodic checks
- Check site analytics during heartbeat rounds
- Alert on traffic spikes — unusual activity significantly above normal baseline
- Alert when new referral sources appear that weren't there before
- Note weekly milestones when they're reached
 
## When to mention analytics unprompted
- Traffic spike detected (2x or more above recent average)
- New referral source discovered
- Weekly milestone reached

Your bot will now proactively mention analytics when something noteworthy happens — without you having to ask every time.

What You Can Ask Your Bot Now

Once everything is set up, just ask naturally in Telegram:

• "How are my analytics?"
• "What's my traffic looking like this week?"
• "Any traffic spikes today?"
• "How many page views did we get yesterday?"

Your bot runs the script and returns a formatted report. Scheduled reports arrive automatically. Heartbeat monitoring alerts you to spikes without you having to check.

Customize What You Track

The Google Analytics Data API can pull almost anything. Extend the script with additional metrics:

Metric	Add to script	What it tells you
Bounce rate	Metric(name="bounceRate")	% of visitors who leave immediately
Session duration	Metric(name="averageSessionDuration")	How long people stay
Country breakdown	Dimension(name="country")	Where your audience is
Traffic source	Dimension(name="sessionSource")	Where visitors come from
Top pages	Dimension(name="pagePath")	Which content performs

Track multiple websites by adding more entries to the PROPERTIES dictionary:

PROPERTIES = {
    "main-site": "123456789",
    "blog": "987654321",
    "landing-page": "456789123",
}

The Bottom Line

This is what AI automation should feel like — tools that actually save you time and keep you informed without extra effort. You don't check dashboards anymore. The data comes to you, formatted, on your schedule, in the chat app you're already using.

The pattern works for anything that produces data: sales reports, email metrics, social analytics, server uptime. Connect the data source, write a script, teach your bot with a skill file, schedule a cron. Thirty minutes of setup, permanent time savings.

---

Tools We Use and Recommend

---

Keywords: OpenClaw Google Analytics, AI bot analytics reporting, Google Analytics automation, OpenClaw skill tutorial, automated analytics reports Telegram, GA4 Python script, OpenClaw cron analytics, AI assistant website traffic, Google Analytics bot integration 2026