http://FreeStrategy.info
Edited by Norbert Bollow

Version 0.1-alpha

February 28, 2003

Abstract:This Net.Book outlines strategies for running a successful business while at the same time making valuable contributions to the Free Software community. The main point is that with an appropriate strategy, significant business benefits can be obtained from the "reputation game" social dynamics which are at the heart of the Free Software community.

Introduction
A free software business is a company which does not exist solely to earn as much money as possible without incurring unacceptable risks; it's a company where contributing to the development and improvement of Free Software, and upholding the principles of the Free Software movement, are among the basic values of the company. (Here the term "Free Software" refers to computer programs where every user has rights to read and modify the source code, and also rights to redistribute the software, with or without changes, as source code or in binary form. The Free Software Foundation maintains a precise definition which can be found at http://www.fsf.org/philosophy/free-sw.html . There are also people who, because of differing views on matters of philosophy, see appendix B, prefer the term "open source".)
Many people seem to think that there must necessarily be serious economic disadvantages for Free Software companies. While it is certainly impossible for a free software business to have a business strategy of creating an illegal monopoly (this is what US courts found Microsoft Inc. to be guilty of) I predict that precisely the absence of monopolies from the realm of Free Software will make Free Software businesses so much more productive than competitors with proprietary products that (with the exception of commercial computer games, see Appendix D) in the long run, Free Software will certainly dominate the information technology market. In fact the only reason why commercial proprietary software is still surviving in most areas (in spite of the competition from Free Software) is that so far most Free Software has been produced by volunteers and not by companies that can afford prefessional usability research and testing.

Publish or Perish
Once a useful program has been released as Free Software, it should be assumed that everyone who wants a copy can get a copy very inexpensively, usually either by downloading it from the internest or by buying a collection of Free Software which contains the desired program. (All current commercial GNU/Linux distribitions contain many useful Free application programs and not just the bare operating system consisting of kernel, compilers, libraries and utilities.)
In this environment, if you develop useful Free Software, you have nothing to gain by not making your program available for gratis download via the internet. If you don't put your software up for download, someone else will, and you have lost the opportunity to use the download website for advertising your commercial offerings.

Selling Free Software
In spite of the above, selling Free Software is still a good idea, even when for most companies this business activity alone will not generate enough revenue to sustain the company. There is a page called "Selling Free Software" on the website of the Free Software Foundation at http://www.fsf.org/philosophy/selling.html that encorages you to charge as much as you wish or can. The following remarks also on that page:
Free programs are sometimes distributed gratis, and sometimes for a substantial price. Often the same program is available in both ways from different places. The program is free regardless of the price, because users have freedom in using it.
Consequently I encourage you to make your programs available in both ways, possibly from different webpages. You can differentitate between your commercial and your non-commercial offering by including a limited warranty with your commercial software distribution. I suggest to let this warranty consist in the promise of fixing bugs that are reported by the customer, up to some per-customer limit on the number of developers hours.
Selling Services
Since selling Free Software (i.e. distributing it for a fee) will in general not generate enough revenue, most Free Software businesses will have to sell services. Some promising areas of business include
Consulting services (empowering companies to run in-house business processes smoothly)
Outsourced business processes (perhaps by means of DotGNU webservices, see below)
Training
Licensing of Commercial Free Software
The GNU General Public License, or GNU GPL for short, is a very good license for commercial Free Software. This is proved for example by the fact Microsoft Inc. has waged a war against this license, warning whoever will listen that the GPL is a "dangerous" license. As Lawrence Lessig, Professor of Law at Stanford Law School writes in his introduction to "Free Software, Free Society: The Selected Essays of Richard M. Stallman", the dangers mentioned by the Microsoft spokespeople are largely illusory. From the perspective of a company like Microsoft Inc. the GNU GPL is however truly dangerous because it is a real threat to the effective monopolies of proprietary software companies. For precisely this reason I recommend using this license as much as possible for software of all kinds, commercial or not. (If you want people to be able to create computer prgrams that combine some code from your program with some code from other another program that has a GPL-incompatible license, this can be achieved through dual-licensing or linking exceptions).
The GPL is so powerful because of a feature called "copyleft" that disallows proprietary software companies from simply taking the Free code and making it part of proprietary programs, thereby reducing the amount of freedom that computer users get.

As long as proprietary software hasn't completely gone out of fashion, this "copyleft" feature of the GPL even allows to make money by selling exceptions from the copyleft terms to company who desire to use the code in proprietary programs. This strategy is used for example by MySQL AB, the company which develops the MySQL database software.

Selling exceptions to the GPL copyleft is possible only when all of the copylefted code in the program has been produced by the same company. Given the general generosity of the Free Software community, the alternative strategy of inviting the community to make GPL'd code contributions to the development project may possibly be more beneficial to your company.

It is not difficult to create new copyleft licenses which are not compatible with the GNU GPL. That however is generally not a good idea, because such incompatibility blocks useful combinations of modules. As the GPL FAQ says, the mere proliferation of different free software licenses is a burden in and of itself.

The distinction between investments and pro-bono contributions
From a business perspective, there are three entirely different aproaches for contributing to the Free Software community. One is to through investments, where the company hopes that the expense will result in business benefits which are big enough to justify the expense. The value of this benefit is often devided by the value of the investment to obtain a quantity called Return On Investment (ROI), and investments should be made only when the expected ROI is significantly higher than 100%.
A second approach is to make commercial offerings where from the beginning a fixed part of the revenue is earmarked for advancing the good cause of Free Software. This may increase your sales as well as benefiting the cause if many of your potential customers have sympaties for the Free Software movement.

Alternatively, and in addition you can make pro-bono contributions where you use a part of your profits to advance Free Software projects which you consider important. This option is generally only available to companies which are very profitable; all other companies need to consider very carefully what investments will allow them to reap a good ROI.

ROI from Research and Development Investments
The benefit from Free Software R&D can come in different forms. If you sell products or services that are directly related to Free Software that you have deloped, you can benefit in two ways. On one hand the Free Software is used in creating the cutomer benefit that the customer ultimately pays for, and on the other way, pointing out that part of the revenue is earmarked for recovering research and development costs of the program will make potential customers more inclined to buy from you than from a competitor who is just using the software without having invested into developing it.
You can also benefit from contributions to the Free Software community that enhance your reputation in the community, this is explained in the section "Difficulties with Business Alliances" below.

Generally speaking, the best ROI can be expected from those research and development projects which have good synergies with your commercial offerings and which in addition give you a good reputation in the Free Software community. Big companies which benefits significantly from having a lot of good Free Software available need to keep in mind that they will only be able to build a good reputation in the Free Software community through making big contributions; this will be expected of them.

The Importance of Business Alliances
At the time of the industrial revolution, starting a successful business usually required a huge amount of capital, and in some areas of business, having a lot of capital is still a key factor. For an industrial company, it is typical to raise capital and use it to buy other companies when the capabilities of those companies turn out to be of great strategic importance.
For Free Software Businesses, it is (or at least should be) impossible to raise the kind of capital that this strategy requires, because there are no "hard assets" that would give the investor real value in exchange for the investment. In addition, acquisitions of companies that create value primarily through intellectual work usually don't work out in practice: The acquiring company tries to install a different corporate culture or at least a different strategic purpose, thereby causing key people to leave; this process destroys much of the value of the acquired company.

In addition there are serious problems with acquisitions of companies in far-away countries with a very different culture. It is quite possible to have control of the majority of the shares of a company without getting much real influence on the real-world actions of the company which should in theory act according to what the majority shareholder says. So the acquiring company may have expended a lot of capital without getting the desired strategic benefit.

For these reasons, Free Software businesses are much better off cooperating with each other instead of trying to control each other through capital acquisition. The term business alliance will be used from now on to describe any form of mutually beneficial cooperation between mutually independant companies. Difficulties with Business Alliances

While business alliances are becoming more and more important in today's economy, but they generally have a high rate of failure (depending on how you count, about half of all such alliances fail) and the negative impact of failed business alliances on the concerned companies is often very severe. Indeed if both partners in the alliance are not really interested in anything but their own profits then it is very easy to see why very often, business alliances fail from lack of trust between the alliance partners. The vast majority of business alliances are never developed to the full extent that would be possible if issues of mistrust wouldn't get in the way.
Having a good reputation in the Free Software community helps with reducing these problems of mistrust. Your company's past and present genuine contributions to the Free Software community (for example by paying developers for working on important free software projects through programming, writing documentation, or usability research) are important to increase the amount of trust that Free Software minded individuals and companies have towards your company, while at the same time your company needs to take care to avoid violating the principles of the Free Software movement. In addition there is a psychological factor: When two people work on contributing to the same Free Software project, it becomes very easy for them to develop a strong interpersonal rapport and a feeling of "we". That also helps to build strong, stable business alliances between their companies.

By the way, here is a reason why I expect business alliances to work better for companies that adhere to the stricter principles of the Free Software philosophy than for those who adopt the more permissive Open Source philosophy, which does not consider software freedom a matter of ethics: In a community with less clear moral standards it is much easier for untrustworthy companies to get away with just about anything while still maintaining a moderately good reputation. On the other hand, when a company builds a good reputation among thoughtful members of the more principled Free Software movement, then the company has a valuable asset that could easily be lost through any action that would be perceived by the Free Software community as a betrayal. It is much easier to trust a company not to commit certain kinds of actions if that company has much more to lose than to win from such actions. This makes it easier to trust principled Free Software companies than companies which subscribe only to the Open Source philosophy.

Another thing to keep in mind when considering business alliances is that alliances with competitors rarely work well. This is not only a trust issue (it is often difficult to trust your competitor) but also a matter of the stability of the win-win relationship. Business alliances are all about mutually beneficial interactions between companies. When two companies have similar offerings, all it takes to turn a win-win arrangement into an unacceptable win-lose situation is a small change of strategy at one of the alliance partners.

DotGNU Webservices
Research shows that besides issues of trust, there is one other major cause of failure for business alliances: the complexity of managing them properly. In today's rapidly-evolving economy, the job of managers is difficult enough already even without the additional complexity that is introduced by business alliances. It is therefore important to reduce this additional complexity as much as possible. Webservices seem to be a promising approach to achieving this; the DotGNU project aims to create a platform for such webservices.
Marketing
This section still needs to be written.
The Problem of "Fair" Prices And Salaries
This section still needs to be written.
Market Research and Business Planning
This section still needs to be written.
Quality Management
This section still needs to be written.
How To Contribute To This Net.Book
Please post contributions to the general discussion about Free Software businesses, as well as proposed new text for this net.book and bug reports to the FSB Strategy mailing list.
Appendix A: On Defending Our Freedom
This section still needs to be written.
Appendix B: Philosophical Differences
Since Richard Stallman's announcement of the GNU project in 1983, there is a Free Software movement which holds that it is ethically and morally wrong to deny software users access to the source code of the programs they use, or to disallow making improvements and sharing them. In 1998 a large group of Free Software users and contributors decided to stop using the term "free software" and say "open source software" instead. While their definition of "open source software" is almost identical to the definition of Free Software, the difference between the philosophy of the open source movement and the philosophy of the Free Software movement is very significant. The Open Source movement and the Free Software movement share the values of generously sharing source code and collabroratively creating community-owned software, but the movements differ with respect to the underlying philosophy. While the Free Software movement is based on principled ethical reasoning, the Open Source movement does not agree with considering the distinction between Free Software and proprietary software to be primary a matter of ethics. For example, one of the business models which Eric S. Raymond recommends in his paper The Magic Cauldron is to "use open-source software to create or maintain a market position for proprietary software that generates a direct revenue stream." This kind of approach is totally unacceptable from a Free Software perspective.
Appendix C: What Scientific Economics Research Says About Free Software Business
This section still needs to be written.
Appendix D: On Commercial Computer Games
This section still needs to be written.
If you like this essay...
If you like this document, you may be interested in Norbert's monthly email letter.
Copying permission
Copyright (c) 2003 Norbert Bollow
Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU Free Documentation License, Version 1.2
or any later version published by the Free Software Foundation;
with "If you like this essay..." being an Invariant Section, with no
Front-Cover Texts, and no Back-Cover Texts.