Knowing it works.
WHAT TO MONITOR
Availability, from outside Error rates Response times Certificate and domain expiry Whether scheduled tasks ran
Business outcomes: form submissions, orders, sign-ups
WHY FROM OUTSIDE
A site cannot reliably report its own unavailability, and can serve an error page while appearing healthy.
WHAT TO CHECK
That a real page loads and contains expected content.
WHY BUSINESS OUTCOMES
Because every technical check can pass while a broken form has stopped all enquiries.
WHAT PRODUCES NO ERRORS AT ALL
A broken form, a failed integration, a silently disabled feature.
WHAT TO ALERT ON
Unavailability Error rates rising sharply Certificates approaching expiry A sharp fall in expected activity
WHERE TO SEND ALERTS
Somewhere independent of the site.
WHAT TO REVIEW
The error log, after every deployment and periodically.
WHAT TO ACT ON
Repeated errors, which indicate something genuinely broken.
WHAT TO CONSIDER
A service collecting errors from browsers, which reports failures users experience.
