In order to assess (and ultimately improve) the reliability of the BM tool and the current test environments, changes reported by the Top Changes feature should be manually verified.

The usefulness of the tool is likely to be low if the majority of stable changes reported turn out to be just false positives that reflect unstable behavior in the test environment rather than changes in the product (Qt).

Note: QTQAINFRA-244 is an example of a regression that was not reproducible on a particular system.