A travel marketplace reports checkout timeouts that backend metrics miss, needs proactive alerts from major regions and real browser latency by page and journey. Which user-experience monitoring design captures synthetic and real-user checkout visibility?

The correct design succeeds because Synthetics canaries run scripted checkout transactions from major regions with alarms on failure and duration when backend metrics miss timeouts, while CloudWatch RUM instruments real browsers for page load, navigation timing and journey analytics with client-side alerting, together delivering proactive and real-user checkout visibility. The Synthetics plus Evidently design fails because Evidently (which AWS has since discontinued in favor of AWS AppConfig) offered feature flags and launch experiments rather than continuous browser telemetry, so synthetic coverage remains but page-level real-user latency and journey breakdowns are lost. The RUM plus X-Ray design fails because X-Ray traces backend service dependencies for requests already received, which improves backend diagnosis but provides no scheduled global probing for proactive regional alerting before users report timeouts. The Synthetics plus Internet Monitor design fails because Internet Monitor measures internet weather and edge-to-region availability, not instrumented browser timing per page and checkout step, so regional network signals remain but page and journey latency stay invisible. Exam caveat is to map keywords browser, page, and journey to RUM and proactive, regional probing to Synthetics rather than choosing backend or network observability. Practical check is to tag RUM pages for cart, checkout and confirmation, then correlate RUM latency spikes with canary alarms by region.