Significance
Much of the recent research on Free/Open Source software (F/OSS) explores why individuals volunteer their time and effort. In fact, not all such motivation is purely personal or altruistic. Many firms are quite happy to have their employees participate in F/OSS development and see this as just another part of the job. Some firms apparently feel that F/OSS may help the bottom line.

This paper explores one powerful reason why this is the case: firms have widely different needs, standardized software products do not meet all of their specific needs, and F/OSS allows skilled users to tailor software to their specific needs. Because the source code for F/OSS programs is available to all users, it becomes, as the authors put it, a "toolkit for user innovation." Skilled users make incremental improvements to programs, which helps the bottom line at their own firms. And when these improvements are made available to others, this helps the bottom line at other firms as well.

The authors find that the needs of these users do indeed vary widely, and that customization helps them to meet these needs. Even though the standard Apache program now comes with a large number of user-requested enhancements (over 6,000 had been incorporated by 1999—see Mockus, Fielding and Herbsleb), most users took advantage of some extra form of customization. Some firms added security modules developed by third parties (33 percent), while others modified the code themselves (19 percent).

As one might expect, the firms that customized the software were more likely to have skilled employees in their server maintenance groups; these firms gave the product substantially higher satisfaction ratings. Although the firms that did not customize gave lower satisfaction ratings, the pattern of their ratings of the product’s various functions followed the pattern of ratings given by the customizing users. The authors see this as evidence of an organic improvement: as skilled users make modifications that are then included in the standard product (an essential aspect of F/OSS), the needs of less skilled users will also be better met.

The authors also estimate users’ willingness to pay for additional improvements to the software. These values appear quite substantial, even after substantial conservative adjustments.

These results suggest that firms motivated by conventional economic concerns may have good reason to embrace F/OSS. The "public good" character of F/OSS does not necessarily mean firms will invest too little in its development, nor are its developers driven solely by altruistic motives.

References

Mockus, Audris, Roy T. Fielding and James Herbsleb, (2000). "A Case Study of Open Source Software Development: the Apache Server"