Ronald Kuetemeier worked for 13 years on an advanced distributed system, from the early days of UNIX networking to Jini lately. Throughout his career at a fortune 500 company Ronald Kuetemeier got involved with strategies by working in what it is normally called a "Think Tank".
For the OSS community Ronald Kuetemeier provides Mist, a distributed Object system, and he is working on a networked spell checker for small or distribute "appliances". Basically providing ispell or aspell services over the network, by using Mist technology. In the future Ronald Kuetemeier might show a system with automatically distributes it self, transparent to the user, when run on networked appliance or stays "monolithic" when used on a "powerful" computer.
This paper is written to highlight some aspects of the software industry and show how to compete in today's closed market by changing the market.
Today we have mainly three processes of developing software and generating revenues from it:
It is clear how to make money from proprietary software. Develop something that fits into a broad market and sell it as shrink-wrapped software. However the limitations become clearer over time. There are only a small number of people that can resolve bugs and its extendibility is very limited for other companies or people who need special functions added to the core functionality. Because of this other firms have specialist in extending the core which sometimes results in an unstable product and finger pointing in the event of problems. Since there is always the risk of new functions appearing in the next release of the original core system (which might replace add-ons) the general support for innovation from the outside is limited to non-existent over a long period of time.
The "open but not open" approach builds upon open protocols and adds proprietary extensions to those protocols to lock the customer in. This business model can be expected to suffer the same problems as the proprietary model. Companies using this model will have to compete directly with companies using the proprietary model.
The "open" part of this business model can be seen as just a marketing point since the customer is confronted with proprietary extensions for which he paid money. Since the source code is not freely available the man power used to build this code must be as good or better than those in companies which build proprietary systems.
While the Open Source Software movement does not suffer from the same limitation to innovation and bug correction as the proprietary or "open but not open" models, it has some apparent problems. There is no revenue stream for its developers at first glance. The only revenue streams I am aware of are services and extended documentation via the publication of a book. Therefore this model is best for companies which provide service. Since new innovation generates new customers who require service for the new functionality, this model is expected to thrive when driven by service companies.
Service Companies driving the OSS model:
The most daunting strategy question for most service companies in the software industry today is, "How do we succeed in a market where the most financially successful company in developing software becomes our competitor by providing service in order to create a new revenue stream?"
The best way might be to create a market in which one revenue stream that supports the competition does not exist. The argument, "We wrote and have the source code therefore we know best how to serve your company" has to be leveled out.
There are two ways to do this.
This seems very compelling for a company with an already established reputation in service and no major revenue stream from products that compete with OSS products. Or from an Internet or Network service provider with no interest in developing shrink-wrapped Software, but with the interest to provide innovative services or applications from the network for computer appliances. It becomes a question of how to keep the OSS movement alive and how to influence the OSS market to build or provide products needed.
Create the market for services:
The first step can be seen as creating the OSS service market. Release software into this market as OSS. Sponsor OSS development for products which are either innovative or needed for a product portfolio. This enables customers to switch over from proprietary software and require service.
The long-term question that arises from this strategy is on how to keep this market alive if its developers have no income. The dependency on good will over a long time period is a risky assumption for a company that needs revenue.
Assuming that the decision is to focus on service revenues and stay out of the direct development, some of the revenues have to go back into development to support future growth. To minimize the investment, the focus here is to create the most effective environments known today in developing software - namely small, focused and highly motivated teams.
This could be achieved through the investment in small "virtual" companies that provide the core team for development and use outside people interested in the product to be developed for documentation and testing. This way, a product portfolio can be built and talented people can be recruited for the service business. The developers not interested in maintenance can move to another project in OSS or be recruited to develop for vertical markets that have to interface with the OSS product. By not releasing the developed product as OSS the interest in "support" can be assumed to vanish since there is no "sensitive" for these people to stay involved. On the other hand there might be a chance of creating a market for freelance writers, who write documentation for OSS products professionally. Creating a revenue stream by publishing a book or creating a Web side and generating revenue from advertising, or simple get paid by a Web portal that provides this service for it's customers.
Most companies buy services from a known source and might also judge on in-house knowledge of the products. Recruiting just the best or "main" developers instead of the whole development team can keep cost down and service competition from becoming too successful.
The consumer market will be most likely move into a specialized appliance market. Services can be sold by providing an easy way to have the OSS already installed. If the processing power is insufficient because the appliance is too small, applications can run on a network and a service fee is paid by either use or support of the application on the network or via advertising.
In any case, OSS provides a way to rapidly create the underpinnings for a service market of scale since the entry into this market is mostly free. This might create a lot of competition for participants at the beginning but it also creates the market of scale for consumers where consumers have a choice of service and are more willing to participate.
At the very least, OSS can be seen as a new start for the software market by leveling out the advantage of existing shrink-wrap product suppliers. Rules can be developed as the market develops. Providing OSS does not only pay software developers in this market, they also have a free distribution channel for innovative products. Innovative companies will require software that supports their core competency as a business organization. Information technology will be used to support this competency. Software companies that have a proven record of innovation and know how to interface with non-core business software will build this software.
If everything works out we are witnessing the birth of a new industry, all over again.