In a fast-moving development organization, with each new build, there’s a plausible risk that the product might have somehow got worse. To address that risk, testers often find themselves responsible for—and overwhelmed by—the development and maintenance of an ever-growing automated regression test suite. Is this the best use of the testers’ time? Are the testers in the ideal position to do this well? What about the other forms of product and project risk—the problems that aren’t related to regression?
Michael Bolton identifies factors to consider when linking the risk of regression to your overall testing strategy.