Determining a relative measure of build quality for a built product based on analysis of historical test data from test cases applied to a series of builds of the built product, by identifying one or more first sets of test results, each first set of test results being derived from the application of a set of test cases to a series of builds of a built product; determining a test quality coefficient for each test case based on the respective test results from the first set of test results; identifying a second set of test results derived from the application of one or more of the test cases to a further build of the built product; applying each test quality coefficient to each corresponding test result from the second set; and calculating the mean of the second set of test results as qualified by their corresponding test quality coefficient.