Uncovering hidden defects

Often, testing can be carried out in parallel, with many parts of the test plan conducted simultaneously. As long as they are independent, the only limit on the number of tests you can run at once is the availability of test systems or testers to carry out the tests. When two tests interact, for instance, by requiring mutually exclusive settings, they either need separate test systems, or you need to run them serially, one after another. Testing serially is much slower, and you should avoid it wherever possible.

Another case where testing has to be run serially is when a bug blocks further testing. That first bug must be fixed before you can run tests and find bugs in the remaining functionality. Any issues that couldn’t be tested are hidden behind the first bug.

For example, if signing up new users doesn’t work, then you won’t be able to find bugs when multiple users sign up simultaneously. The bugs with multiple users are hidden...