Two dimensions of software acquisition

New field data supports the contention that two dimensions of the software acquisition problem-cus-tom/package and insource/out-source-are interrelated at several levels. S OFTWARE acquisition is a critical concern for firms worldwide, and its impact on their business processes is expected to grow. U.S. firms spend more than $250 billion annually acquiring software [18]. Software investments can result in substantial productivity gains and strategic advantages [2], but not all such investments are successful by any means [3]. To realize these benefits, firms must identify and understand the factors that most affect their software acquisition decisions. In this article, we develop a general cost-benefit decision framework for the software acquisition problem. We use this framework and software acquisition data to investigate six hypotheses characterizing software acquisition. Our framework and empirical results offer managers a basis for structuring and benchmarking their software acquisition decisions and offer vendors insights into the characteristics of the software projects that would be most beneficial to target. The software acquisition problem is depicted as two-dimensional (2D)-what acquisition approach to use (custom develop the software or base it on a package) and who should complete the task (internal resources or external providers). While aspects of the sourcing decision have generated frequent study, the custom/package decision has not previously received attention. In addition, our decision framework captures the major tradeoffs a firm faces when making these two decisions. Previous theoretical studies have focused on singular aspects of the insource/outsource decision. Our framework unites these aspects. It builds upon the contracting and monitoring cost issues in the transaction-cost and incomplete-contracting literatures and includes the tradeoff between monitoring and production costs developed in the agency theory literature [13, 17]. The model also encompasses the various managerial rationales provided in the extensive empirical literature regarding firms' sourcing decisions for information technology (IT) services [8, 12] and the limited literature on software acquisition [16]. Our analysis extends this literature by using firm actions, not post-decision rationalizations, to determine the key factors behind acquisition decisions. Following findings in the marketing literature [5], relative factor weights inferred from actual decision data should provide greater explanatory power than previous studies have offered. The 2D, four-option sourcing decision is depicted in Table 1. Using packaged software involves identifying the company's information processing requirements, evaluating alternative packages, selecting one, possi-Software Acquisition