The Business of Free Software: does a model exist?

A recent article on ZDNet titled �The commoditization of software� By John Carroll [URL: http://zdnet.com.com/2100-1107-992824.html] opines that:

�Commoditization is a powerful force for consumer benefit. In hardware, however, there was never any risk of free hardware pulling the economic rug from under the industry. Commoditization, taken to the point where software is practically free, would run contrary to the interests of software developers who have an interest in earning a living from their craft. In short, a market where the price of software plummets too low will rapidly drain the value of programmers, and hence cost many programmers their jobs and drain the industry of investment as investors chase more lucrative places to put their money.�

In view of a thematically similar posting of the AsiaOSC site [URL: http://www.asiaosc.org] about how cost should not be used as a bargaining chip in the promotion of Free and OpenSource software, the media focus has shifted to the inherent economic model or lack thereof of the movement. I had earlier on, at the AsiaOSC site posted a reply to the effect that when FLOSS proposals for academic institutions are carried out, TCO is one of the major factors for consideration. If more stability, industry standard solutions are available at a lesser investment, then in the end cost does play a major role in the decision making process. The content and the reply is available at the AsiaOSC site as well as my personal weblog Random Thoughts [URL: http://sankarshan.blogspot.com]

It is a sad fact of the economy driven by FMCG media spends that �free� as a literal meaning has come to be associated with something obtained without paying currency. Notwithstanding the rallying cry of the Free Software Foundation [www.fsf.org] �free as in speech, not as in beer�. Free means gratis. End of story. FLOSS activists have repeatedly run into such obstacles at project proposal stages, review and implementation stages. And thus, a business model needs to be visualized for the effort. While the quote attributed to RMS has not being made available in the form of an entire quote, given RMS�s current vision to shift the end goal of programmers to a utilitarian level by changing their needs axis, it is not quite difficult to guess what he was talking about.

John states that �. . . . the vast majority of open source software is, in fact, available at no cost.� � very true. And this is perhaps been one of the driving and enabling forces of adoption of technologies and toolsets provided under the umbrella of FLOSS. The community model of development has ensured that successive contributions to FLOSS projects can be made available globally from a single point of reference � the repository. But what good are these products by themselves? The products taken in conjunction with solution sets and solution implementations are the main driving engines of growth. He further compares Adobe�s bottomline and Photoshop�s piggybacking on the Adobe engine based on price. Good point, but are we not comparing apples and peaches? In both the cases, the software houses are competing with price differential and a value addition second. And to take on one of his examples, in India, Oracle, which has a huge database penetration, is promoting the GNU/Linux version of its current DB suite as the stable and industry standard platform. IBM cannot be clubbed in this group simply because being the earliest of the behemoths it has diversified into so many areas that one corporate policy identification is probably difficult.

Profit earnings in the software segment are not like the consumer durables industry where in case topline growth remains stagnant, business process restructuring can ensure some amount of bottomline ego boost. Case in point being the FMCG giant HLL. Sales and extraneous factors directly affect revenue stream. Thus, a revenue model must factor in periods of sluggish sales as well as ramp up costs for investing in new technology. However, what is most important is that a revenue model must also realize the value-addition that software should provide to a business. �Free software movement� is not about preaching/propagating that �software should not be worth any monetary value�. Free software is about giving the users and the technicians a choice. Freedom to choose is one of the major drivers in any technology oriented society. To compare the freedom and the breadth of such freedom, let us for a moment sidestep and analyze the effects of the Shared Source Initiative by MSFT. Considering that the customer is a partner in the initiative [as it has signed the agreement], hypothetically let us assume that a mission critical application is proving to be troublesome. Burning the midnight oil, the team finds out the cause of the problem, patches through a solution [thoroughly cost effective] that involves some little modification to the Windows code. Can you implement it? Most certainly not!! The Shared Source license permits the partner to look and not modify. Consider the same under GNU/GPL and imagine is the mission critical application were to be the revenue bulwark of the company. Free software allows changing the code with respect to the hardware; proprietary software makes it mandatory to change the hardware to fit the code.

Software valuation is 2 fold � first what the developer perceives as the worth of his/her coding effort and second what the market perceives it to be. In the Indian context, with the artificially hiked prices of software suites, the market value of any software is skewed. Consider this fact, the Borland Turbo C++ suite, required in most educational institutions, is available for INR 8800 for a single user license. While compiler environments like Bloodshed DevC++ [released under GNU/GPL exist], rampant usage and proliferation of unlicensed software has ensured that innumerable copies of such environments are used in schools and colleges not to mention student desktops. Price rationalization is one of the biggest factors that creates divide in �developing countries�. It is absurd for such economic models to pay what is the price for the piece of software in other �developed� nations. This is just one of the many examples that can be put forth to justify the free software model. John�s hypothesis that pricing it low and/or promoting it as free might just make the pricing structure head south is, I feel, unfounded. Enough business case studies have been carried out to identify Free Software business models. The earliest and path breaking work being Eric Raymond�s �The Magic Cauldron�. Models exist that make money from free software usage and providing solution sets, and companies exist that are doing nicely in a sluggish economy based on these models. Moreover, market economics has an aversion to anything being priced too low. What however is the fear is that a �low price� might just be equated with �inferior quality� of software. Taking the example of Redhat itself, it is absolutely based on a service provider model. Even though the GNU/Linux ISOs can always be downloaded from any of the ftp sites, the money that the company charges for the box packs is about the value addition it brings to the client end in terms of support and service. The companies and individuals will not pay more for customization, but will pay for widgets and toolkits that add value. Value addition and estimation of such value using conventional toolsets will lead to more implementation. Free software offers the advantage of providing a veritable array of off-the-shelf toolkits that could be mixed and matched to provide a working solution in the least possible time. Sort of like the �Lego� of our earlier years.

The software segment is akin to the service industry with value addition and ROI being brandished at each stage as the main plank of IT spend. And Redhat managed a �thin� profit while selling the product, so why should investors [potential and current] shy away from the company? Mandrake�s predicaments were of its own making. With corporate giants like Sun, IBM, HP gunning for GNU/Linux, sometimes at the cost of their own product chains, it is very much possible that the market would be looking at a shakeout in the number of competitors involved. Certainly it might resemble a gunfight as in the old westerns, or then it might also lead to consolidation efforts like that of UnitedLinux. But a failed business model? Never !!

Sankarshan Mukhopadhyay is a Free software enthusiast and a member of the iLUG-Kolkata chapter URL: www.ilug-cal.org He can be reached at sankarshanm@softhome.net


- posted by sankarshan @ 3/21/2003 06:22:00 AM