OPEN SOURCE: Going Vertical

The next major wave in the open source movement is vertical market applications, as Eric Raymond pointed out in The Magic Cauldron. There is a fundamental flaw in the way software value is perceived. Software is not a factory produced, assembly line product . Only the packaging is. Software is just as much art, as it is science. If this were not so, there would only be one accounting program or one video store management application. When an application is selected, it must perform the necessary tasks (the science), but it must also appeal to the customer in other ways that relate more to how comfortable they are with the user interface and the process flow (the art).

It is therefore this combination of art and science that gives software its inherent use value. More often than not, a software package is chosen based more on art than science. This can be seen in the popularity of some office suites, where the slick user interface and familiarity overcome the underlying technical weaknesses to build huge market share. The same premise stands for vertical market applications. Those that imitate a familiar interface stand a better chance of being accepted in the marketplace simply because they seem easier to use, even though they may suffer serious technical weaknesses.

Taking an existing vertical market application open source is risky. A company considering this move will be concerned that their hard-earned intellectual property is now available for use by all competitors. Even worse, it may be useable to competitors that didn't even exist before the release. Therefore, they must develop a solid plan to not only maintain but increase their market share in a short period of time. This plan must include active management of the open source project and some type of compensation sufficient to capture quality developers as well as an active user community. Why? Simply because we do not live in the utopia of a total gift culture.

The release plan requires some crystal ball gazing, as well as emotional detachment from the existing application. Some of the questions to be asked are:

1. What was the time to market?
2. What significant changes in technology occurred during development time that could improve the application if it was started from scratch today?
3. How can we provide a migration path from the existing application to the new open source application?

Many open source vertical market applications are actually started as either in-house or hobby projects. In these cases, one of the most difficult things to do is create the required emotional detachment. When an individual or small group develop an application, it is built from their view of the world. So the art of the application looks good to them and to the relatively small group that share their perspective. Even though the code suffers peer review and is more often than not technically superior to a closed source equivalent, it may not appeal to a large enough market segment. An application in this state will not likely draw a large developer base.

The successful open source vertical market applications of the future will be built from the ground up as an open source project based on a loose set of requirements. A project basically built around a goal with just enough specifics to get conversations started. The project coordinator has to have a special set of skills that encourage participation by a wide range of people, as well as a willingness to leave things alone when they are progressing well. There has to be an incentive for participation as well. In the case of FreePM, there are different incentives for different participants. Primarily for the physicians and other health care providers, the incentive is to help build the best practice management system available at an affordable cost, concentrating, for the most part, on fixing what's wrong with the current offerings. Since it is a tool for them, it does not affect their market share in any direct way. For the developers on the team, the incentive can vary from skill building to business building. By being a part of the development, they will have intimate knowledge of how the application works, what its limitations are and how it can be extended and customized to fit any given situation. So, they put themselves into a position of being a market leader in physician practice management system installations.

The previous two paragraphs are in direct contradiction to Eric Raymond's preconditions of the bazaar. The difference is in the fact that Eric's perspective (in my opinion) was based on users and developers being the same participants. With a vertical market project, the users and developers must to be brought together on equal footing, at ground zero.

The public launch of FreePM was 28 June 1999, based on this model. Granted, there is at least a years' worth of planning and examination of other projects behind this model. However, it is working well for us and I believe it will work for your next vertical market application.

--------------------------------------------------------------------------------
Tim Cook is currently working in the IT health care community on a variety of projects. The Free Practice Management project can be reached at FreePM.org. There is still plenty of room to join a talented group of health care and applications professionals on this project.