URL monitoring is binary: your URL either responds with an expected status code or it doesn't. It's a necessary baseline, but it has a fundamental blind spot.
A site can return HTTP 200 on its homepage while the login page is broken, the checkout process is throwing a JavaScript error, or the API endpoint powering the search function returns malformed JSON. None of these show up in a URL monitor.
Transaction monitoring fills this gap.
What Transaction Monitoring Actually Tests
A transaction monitor simulates a real user walking through a multi-step flow:
- Load the homepage — verify it renders
- Click the login button — verify the form loads
- Submit test credentials — verify successful login
- Navigate to a product page — verify the page loads with correct content
- Add to cart — verify the cart updates
- Proceed to checkout — verify the checkout form loads
- Verify the payment form renders — stop before charging anything
If any step in this sequence fails, the monitor fires an alert immediately. You know that your login flow is broken before a single user reports it.
Real Failures That URL Monitoring Misses
JavaScript errors in client-side rendering: Single-page apps (React, Vue, Angular) often return 200 OK at the URL level even when the app has crashed client-side. A transaction monitor that tries to interact with the UI will catch this; a URL monitor won't.
Third-party API failures: If your checkout flow calls a payment processor API that starts returning errors, your site's HTML loads fine but payments stop working. A transaction monitor can verify that the payment form renders and submits correctly.
Database query failures: A product page might load its HTML shell correctly (200 OK) while a database query failure means the actual product data is missing. A user sees an empty page. Your URL monitor sees a success.
Broken form submissions: Contact forms, lead capture, account registration — these are critical business flows that a homepage check completely misses.
When You Need Transaction Monitoring
You don't need transaction monitoring for every site. A simple brochure site with no user interactions beyond contact forms doesn't need multi-step synthetic tests.
You likely need it if:
- Your site has a login/authentication flow that real users depend on
- You run an e-commerce store with a checkout process
- Your SaaS product has a critical onboarding flow
- You have API endpoints that backend applications call in sequence
- Your site uses a headless CMS or SPAs where the HTML shell and the content are loaded separately
Transaction Monitoring on ndelog
Transaction monitoring is available from the Warden plan ($7/mo) with up to 10 transaction monitors. Pamong includes 50.
Setup is done through the ndelog dashboard where you define the steps: URLs to navigate to, form fields to fill, elements to verify, content to assert. No scripting required for standard flows.
Alerts go to the same channels as your URL monitors — WhatsApp, Telegram, Slack, or email.
Combining URL and Transaction Monitoring
The practical approach for most developers:
- URL monitors for all URLs — fast and cheap, catches server-level failures immediately
- Transaction monitors for 2–3 critical flows per application — slower to set up but catches application-level failures
A client's e-commerce site might have 1 URL monitor for the homepage and 1 transaction monitor for the checkout flow. That covers both infrastructure failures and application logic failures with 2 monitors.