On event days, a stadium wants synthetic web tests to replace Endpoint Agent data collected from phones in the stands. What should the assurance lead say?
Select an answer to reveal the explanation.
Short Explanation
A rehearsal on an empty stage does not tell you how the crowd will sing along. You keep synthetic tests and Endpoint Agent views because they answer different questions, and a script cannot stand in for phones in the stands. A synthetic pass is not a census of every fan, a capture of the ticketing database, or a replacement for a real phone's browser waterfall.
Full Explanation
A synthetic web test runs a script from an agent and answers whether that scripted path works. An Endpoint Agent on a phone in the stands answers what that real device experienced on event day, including the local network the script may never sit on. Those are different questions, so the script cannot replace the endpoint data. Treating a synthetic pass as proof that every phone had the same experience turns one script into a census of the crowd. Treating the synthetic test as a packet capture of the ticketing database confuses a web transaction with a copy of database activity the test does not record. Letting the script replace a browser waterfall from a real phone discards the object-level timing that phone collected and still fails to represent other phones. Keep both views. Use the synthetic test for the scripted journey from its agent, and use Endpoint Agent results when the question is the experience in the stands. Exam caveat: synthetic coverage and real-user endpoint coverage are complementary, and neither one deletes the need for the other. Operational check: place one event-day Endpoint Agent result next to the synthetic result and confirm they are reported as separate evidence.