What this proves
A healthy page can still contain a broken journey
Both runs loaded the site and configured the product. Only the full submission exposed the later failure.
Website monitoring · State changed
The server still responded and the configurator still worked. Only the final enquiry failed. Repeating the complete journey caught what a simple uptime monitor would miss.
The later run reported: “Project service unavailable. The enquiry could not be recorded.”
Paired runs recorded · Pass changed to fail
The exact repeated test
“Configure a Linea 240 filling system for 6000 units per hour using PET containers. Include installation and operator training. Calculate the fit, then request a consultation for Thomas Becker at Becker Getränke GmbH using thomas@example.test. Pass only when ‘Projektanfrage erfasst’ and reference ‘KW-L240-6000’ are visibly confirmed. If the project request cannot be submitted, report the observed blocker instead of treating the configured form as success.”
Broken form monitoring
The expected confirmation and reference KW-L240-6000 were visible.
Run again
The page stayed online, but its visible error proved the business outcome had failed.
Operational use
Mira can run a saved website test every chosen number of hours or days. Email notifications are triggered when the run health needs attention and when it recovers.
Keep the instruction and visible success condition together.
The same goal is re-evaluated in a fresh browser run.
The email points to a failed or recovered workflow and its evidence.
Evidence note: the film combines two controlled recorded runs to show the pass-to-fail signal. It does not depict the scheduler screen or an email inbox.
What this proves
Both runs loaded the site and configured the product. Only the full submission exposed the later failure.
What it does not prove
The failed service is an intentional variant of a fictional website. It demonstrates detection behavior without claiming a real customer incident.
The film first shows a successful project enquiry and states that yesterday’s successful test does not protect today’s workflow.
The same configuration runs again. This time, the final submission shows “Projektservice nicht erreichbar. Die Anfrage konnte nicht erfasst werden.” The red proof card states that the problem was detected before customers reported it.
Let the successful runs stay quiet. Get the evidence when the outcome changes.