Home » Workflow Automation » Website Monitoring

How to Automate Website Monitoring and Alerts

A website monitoring workflow checks your site on a regular schedule and alerts you when something goes wrong. It can verify that your pages load correctly, check for specific content that should be present, monitor response times, and detect changes that might indicate a problem. When an issue is found, the workflow sends an immediate SMS and email alert so you can respond before customers notice.

What to Monitor

Website monitoring can range from simple uptime checks to detailed content verification. The most common checks include:

Uptime: Is the site responding at all? A request that times out or returns a server error means the site is down.
Response time: How long does the page take to load? If response time suddenly jumps from 500ms to 5 seconds, something is wrong even if the site is technically "up."
Content verification: Does the page contain expected content? If your homepage should say "Welcome" and the response does not contain that text, the page might be showing an error or serving the wrong content.
SSL certificate: Is the certificate valid and not expiring soon? An expired certificate blocks visitors with security warnings.
Specific endpoints: Do your API endpoints return expected responses? If an API returns errors or unexpected data, your integrations may be broken.

Step-by-Step Setup

Step 1: Create a scheduled workflow.
Set the workflow to run every 5-15 minutes for critical sites, or every hour for less critical pages. More frequent checks catch problems faster but use more credits. For most business websites, a 15-minute interval catches outages before they cause significant impact.

Step 2: Make the HTTP request.
Add a block that sends an HTTP request to your website URL using the external API connector. Store the response status code, response time, and response body in variables. A healthy website returns a 200 status code. Anything in the 400 or 500 range indicates a problem.

Step 3: Check the response.
Add condition blocks that evaluate the response. Check the status code first: anything other than 200 triggers the alert path. Then check the response time: if it exceeds your threshold (for example, 3 seconds), flag it as a performance issue. Optionally, check the response body for expected content to catch cases where the server responds with a 200 but serves an error page.

Step 4: Send alerts when issues are detected.
If any check fails, send an SMS alert immediately to the person responsible for the site. SMS is critical here because email might not be checked fast enough. Include what failed (down, slow, or wrong content), the URL, the timestamp, and the specific error. Also send an email with full details including the response code and response body excerpt for debugging.

Step 5: Track status history.
Log every check result (success or failure) to your database with the timestamp, response code, and response time. This creates a history you can analyze for patterns. A weekly report workflow can summarize uptime percentage, average response time, and any incidents.

Preventing Alert Fatigue

A monitoring workflow that fires alerts on every single failure will overwhelm you during a real outage (you get a new alert every 15 minutes for hours) and create false alarms from temporary network glitches. Add these safeguards:

Retry before alerting: If the first check fails, wait 30 seconds and try again before sending an alert. Many failures are momentary network issues that resolve themselves.
Cooldown period: After sending an alert, do not send another for the same issue for at least 30 minutes. Track the "last alert sent" timestamp and skip subsequent alerts within the cooldown window.
Recovery notification: When the site comes back up after an outage, send a "recovered" message so you know the issue is resolved without having to check manually.

Monitoring Multiple Pages

To monitor several URLs, store them in a database table and have the workflow loop through each one. Each URL can have its own check criteria (some might need content verification, others just uptime). The loop processes each URL, runs the appropriate checks, and collects any failures into a single alert message rather than sending separate alerts for each URL. See How to Loop Through Data in a Workflow.

AI-Powered Analysis

For a more advanced setup, add an AI step that analyzes the monitoring data weekly. Send the complete history of response times and errors to an AI model and ask it to identify trends, predict potential issues, and suggest improvements. The AI might notice that response times increase every day at 2 PM (suggesting a scheduled job is competing for resources) or that a specific page has been getting progressively slower over the past month.

Cost note: A monitoring workflow running every 15 minutes costs approximately 4-6 credits per check (HTTP request, condition checks, database log). That is roughly 400-600 credits per day for one URL, or about 12,000-18,000 credits per month. Monitoring 5 URLs at 15-minute intervals costs 60,000-90,000 credits per month. For less critical monitoring, hourly checks reduce costs by 75%.

Know when your website has problems before your customers do. Set up monitoring in minutes.

Get Started Free

View the Chain Commands App