One of the great success stories of the computer age has been Microsoft’s Windows, the dominant PC operating system. Many businesses also rely on another Microsoft desktop software package called Office, a “suite” of programs that includes Word, Excel, and Access, and starting with Office 2000, Internet-friendly features. Office promises speech-recognition software in the future that may make some keyboard and mouse features redundant. In spite of these features, Microsoft has had to work harder at promoting each new version of the software package. The reason is that many users of Office find that its programs are so packed with features that they cannot exhaust all of the possibilities. According to some estimates, even enterprising users with lots of time on their hands are unlikely to use more than a quarter of the features of Word. Consequently, a large number of users are concluding that upgrading to bigger, more sophisticated versions of the Office suite does little to enhance their productivity. Evidently, users of Office are confronting diminishing returns from adding further to the extensive software code stored on their hard drives.

Under what circumstances could adding computer software yield a negative marginal product, at least within a period following the installation of the software?