The Love of Linux Chapter 11: The Business Model
How do you make money on something that is basically free? That is the question that I have been asked almost from the first day that I became involved with Linux over twelve years ago. I believe the answer is becoming more obvious, but it was not when we were involved in Linux. A recent book entitled “Free: The Future of a Radical Price” by Chris Anderson provides a much broader history of the use of free and its future. Chris points out that with processing power and storage more than doubling every 18 months and bandwidth even more rapidly, the pressure toward everything digital being made available for free will be significant. Chris lists numerous examples of products and services that are moving or started free such as the gaming industry. Games have been traditionally a product centric business but the new games are available for free online and Chris predicts that nearly all games will so follow. They are far more profitable than the traditional product business, because there is almost no cost to distribute, they get direct access to those who are their target market, and they can make updates and enhancements for very little cost. They sell advertising in incredibly creative ways. Billboards or posters on buildings, clothes or other product endorsements are being included in the game, etc. Chris also does a paper napkin approach to determine how much money is being made in and around free software and came up with a $360 Billion dollar industry. This and the fact that Chris makes the statement that all things digital will be under pressure to be free, a little more than 10 years after we started with open source and Linux is both almost unbelievable but also very vindicating.
When we first started Linux, we felt open, free access and price, was the wave of the future but at the time consumers were not paying more for bottled water than for soda at the store. Nor had Google become a global power providing a free service by selling advertising. For us, finding the answer to the question of making money on Linux was truly a journey in faith.
In the technology sector, Linux and Open Source really pushed the envelope of free and in many ways paved the way for Google’s success by fueling the Internet infrastructure in an economical way. There were many times I felt like the young man left in a room filled with horse manure. Several hours latter when they returned they found him still digging through the manure. When asked what he was doing, he said, “With all this horse manure, there has to be a pony in here somewhere.” In this chapter, I will try to reveal the pony or ‘show you the money.’ But maybe more importantly, I will share some conclusions that have far greater application then just to Linux or to even the emerging free digital economy described by Chris Anderson. In this new world of the Internet with so much information, nearly every industry is feeling like they cannot keep up on the ever increasing tread mill of innovation. If they invest time and money into producing a product, they often do not have the product cycle long enough to get a return before a competitor has produced something better, faster or cheaper. It is a proven fact that nearly every major invention made was made by more than one person independently in totally isolated parts of the world. With the abundance of information made available from the Internet, inventions will continue at a pace unparalleled in history. There is a principle called the 100th monkey principle derived from a study made on isolated troops of monkeys. These troops had no way to communicate with each other, but almost simultaneously, each troop developed new eating behaviors. Apparently, the monkeys would not sweat potatoes that fell to the ground and got dirty. They would just get themselves another sweat potato. Then, usually a young male monkey in each troop decided to take the sweat potatoes down to the river and wash it off and eat the rest of it. Soon it became an accepted practice in all the troops. Some have questioned the “rapid adoption by unrelated troops” as not being conclusive in the original studies, but similar phenomena has occurred in humans.
Traditionally, the method of protecting ideas, inventions or what is called Intellectual Property (IP) was to get a patent. Today, often patents are primarily there to protect the investors. Historically it was a race to get to the patent office first. Now that is complicated because it is a global economy and it is not sufficient to protect your product in one market, you must register it in all major markets. Also, with the level of innovation that is taking place around the world, how can the patent office keep straight what is a new invention or if it even qualifies for a patent? I read an article that the patent office was going to experiment with a Wiki model for registering patents. Clearly, something has to change because there are so many patents that should not qualify to be patents. Even with a patent, it does not guarantee that you can enforce it. I am not against patents, but they are becoming harder and harder to enforce and more and more expensive to protect. A company can choose to spend its life in a court room making attorneys rich or they can learn some principles from watching the open source model mature. Some can argue that no one is really making money on open source. Clearly there are some examples where open source companies are doing relatively well, but none of them are extremely successful financially. However, there are many like Google that are making billions on free. In some ways, the same forces that make it difficult to make money on open source apply to nearly every other product in today’s market as I mention. The affects of technology innovation and the Internets explosion of information are fueling innovation. So as open source matures and solves some of these concerns, there will be a lot to again learn from Linux, the new Internet economy and the emerging digitally free models. In many ways, the Internet and open source are not only causing significant change in how we do business, but it is like a magnifying glass, simply exposing things that have already changed but we have not been able to see the signs as clearly and as quickly as before. This free exchange and broad availability of data is allowing us to see change quicker. The phenomena of “Who Move the Cheese” is heightened. I am reminded of a statement that Ray Noorda was quite fond of saying. Ray said that a company can fight change and die, accept change and possibly survive, create change and thrive. No one has a monopoly on good ideas. And with free flow of information and the ability to collaborate on projects around the world, innovation is going to continue at an amazing pace. So how does one protect their investment and bring something to market that they can reap a return on?
Caldera experimented with how to make money from open source software from the day that we were incorporated. In some areas, we helped pioneer many of these models in the Linux space and as a result have taken a lot of heat from the community. These models include:
1) Sell services not software (We did not believe this model could sustain itself)
2) Add proprietary value above Linux or the sale of a related product (Give away the razor and sell the blades)
3) Technology innovation (Hold and then release) or hold some features to create a “premium edition” for fee but release openly over time
4) Collaborative cost sharing with specialization — Customer focused solutions
5) Certified platforms/Subscription services
Something can be learned from each of these models and several of them or their combinations have proven successful. Each of these concepts is worth discussing:
Sell Services Not Software
One could argue that the “distribution” business was nothing more than a “service” since none of the companies who offered Linux owned anything. They were just packaging the software. The service only model is most conducive to the open source mentality and has always be accepted and promoted by the Linux community and industry. Caldera was not a pioneer of this model. Frankly, Cygnus and later Red Hat were the companies that are held up as companies who have been successful at this model. It is interesting that Cygnus, even before being bought by Red Hat were experimenting with other ways to make money. I listened to the CTO of Red Hat when he worked for Cygnus give several presentations and sat on several panels with him in which he justified moving to proprietary software because they needed more revenue potential. Red Hat is currently experimenting with a combination of the methods I am discussing with their advanced service product. The advanced server is a combination of technology innovation and certified platforms which I will discuss later. The point is, a pure services model is difficult to build a long lasting, profitable business on. Caldera did not heavily pursue this method because there are two major challenges with this model. First, it is difficult to scale. Services require people to provide them. A person can only physically handle a certain number of accounts or projects at one time. Consequently, to grow the business requires the company to increase overhead. Clearly, there are very successful service companies, but they are not as profitable and consequently as attractive to investors. The margins are much lower. There is one notable exception to the services model that may very well scale, the subscription model which I will also discuss later. The other challenge with selling services only, is that it is much harder to differentiate yourself from your competitors.
Service is about being where your customer needs you to be. It is difficult to compete in services against an IBM Global services, for example, because they are every where the customer needs them to be. You can target services to companies that fall below IBM’s targeted customer size or other requirements, but that greatly reduces you market potential. The recent announcement by Oracle that they intend to provide better support for Red Hat than Red Hat can on its own product illustrates this challenge. The original “distribution service” that started Linux is a good example of the inability of this model to be self sustaining. Cygnus and Red Hat both began utilizing this model because it is highly compatible with the open source mentality and thereby helped them establish market share. Of course, in the beginning, I think Red Hat made a lot more money on selling hats, shirts and other trinkets then on pure services with the exception of the distribution itself. It is important to point out that both Cygnus and Red Hat are held up as icons of successful open source, service based software companies and yet both Cygnus before they were acquired and Red Hat now, are experimenting with other business models to continue their growth. Selling services are an important extension of a business, but they will not be the main source of the business model with the exception of a “subscription” based solution.
Sell Value add above Linux or a related product
Caldera was the first to try a value added concept. In other words, add something proprietary above the Linux kernel and associated packages that the customer was willing to pay for. The Caldera Network Desktop, our first product, introduced a fully graphical desktop to Linux derived from the proprietary Visix technology. As described previously, Caldera spent millions of dollars licensing technology to move Linux forward and provide a reason for buying Caldera’s offering. The only one that really benefited from this effort as also mentioned, were the companies like Netscape, Star Division, Corel and others who owned the technology. Caldera was constantly accused of trying to take something from Linux when in reality; we spent far more on marketing Linux and licensing these technologies and received relatively little return. The problem with adding proprietary elements to open source is that it then clouds the license usage issue. One of the reasons the customer wants to buy open source, is that it is “free.” They want to be able to copy the software onto other systems. Having a license that permits copying and redistribution of software for some pieces and restricts others, simple confuses the issue. Customers will simply choose the less confusing solution even if it may not be as good. It only has to be good enough just ask Microsoft. As complete solutions are created that target specific customer segments, this may become more acceptable. The value-add model is however, the model most large companies like Dell, Hewlett Packard, Oracle and IBM are using. They do not make money on Linux; they simple sell more hardware, software or support. This model works fine for companies who have an established market share in either hardware of software products that benefit from having Linux present. In fact, the Linux model is ideal for companies like IBM who sell a variety of hardware, software and services. The challenge for most Linux companies was they did not have other products that can be sold as a result a free operating system.
The challenge with the value added model is that eventually the concerns of the product making money will dictate the direction of the operating system that is free. A good example of this is the “Unbreakable Linux” offered by Oracle. They have modified Linux so that their particular software will run faster and scale. Red Hat and SuSE (Formerly UnitedLinux) have agreed to integrate the necessary changes back into their respective products. This offers Red Hat and SuSE a chance to have an additional value proposition, the technology innovation advantage, for a period of time, but the agenda is set by the Oracle database. I am not suggesting that this is necessary bad, but what about the needs of the other database vendors? Does that relegate Linux to being a niche, customizable operating system platform? The real benefactor of this approach is Oracle and Oracle’s customers who chose to deploy on Linux. As mentioned before in an article I wrote several years ago, eventually, Linux will be driven by the needs of the hardware and software products that are making money. Another way to explain why this is potentially harmful to the best interests of Linux is to look closely at how the companies like Intel, IBM and Hewlett Packard market and position their Linux products and offerings. They do not position them to go after any Microsoft customers. They are targeting UNIX customers. Customers they do not have on Intel platforms. The one exception to that is IBM, because I believe they are were still smarting from losing to Microsoft on OS2. IBM appeared to market Linux as an alternative to Microsoft in some areas. At least, some divisions were not afraid of talking about it. These examples point out that if Linux makes money, it can begins to write its own agenda and will have broader market acceptance. If the company who is selling Linux is making more money on Microsoft products, they will not promote Linux as an alternative to Microsoft because they do not want to loose the revenue or the potential relationship with Microsoft. (This is why Microsoft carries such a big stick and it is so hard to pin them down on misdeeds. Many companies exclude Microsoft’s competitor’s offerings just because they fear Microsoft retaliation.) As I said before, this is not just about Microsoft.
What brings in the money, will have the mindshare of all those who are part of the distribution system. Therefore, if Linux, or any open source product, wants to be something other than a niche product, it must find a way to be financially self supporting. What I mean by financially self supporting, is that it must be able to not only cover development costs which are lower on Linux because of the collaborative development model, but all the cost to promote and support the specific customer solutions provided by that technology which are as much or maybe even more than more conventional software products because they are so different. These marketing costs require hard dollars so Linux companies must make hard profits. Consequently, while the value added model works well for companies like Oracle, IBM, Hewlett Packard and others because they have proprietary products and services who make incremental sales through promoting Linux, it does not work for Linux itself of for companies providing Linux. As I mentioned, Caldera was the first to try the value added model and we continued to experiment with it up until I left with products like Volution, a comprehensive management system for Linux. The challenge for a start up company with scarce resources is that it becomes difficult to decide on where those resources go. When you must invest resources into developing the operating system or core business, it is difficult to invest in the value-added or new business opportunities. This is difficult for businesses in all stages if they do not have a clear strategy or have lost touch with their customers. Often in mature businesses with products that have become cash cows, management often struggle where to put scarce resources and invest properly. This was true of SCO before we purchased them and apparently true for SCO’s new management and board as they seem to have gone back to what they think will bring them the most money, choosing UNIX over Linux. The value-add or new business model is particularly hard in start up mode because there is no established revenue base and you are just trying to stay alive. Consequently, for most Linux companies, a value-add model was very difficult. Companies like HP, IBM and Oracle could easily play this game because they have the base.
Technology Innovation (Hold and then release) or create a “premier edition” for a fee or a “Freemium”
Caldera also tried technology innovation, the hold and release proposition. At one time, Caldera felt that it could innovate and on the first release hold the technology proprietary and then subsequently release it under a social contract. Caldera’s innovative graphical installer, Lizard was just one attempt at this business model. Other companies like SuSE also experimented with this model with the drivers. The Linux community was suspicious of the company’s intents and usually, all that resulted was to launch an all out effort by the competitors to develop the same functionality as you have. This resulted in multiple ways of doing the same thing, duplication of effort, and fragmenting rather than creating standards. If your competitor happened to open source the technology before you did, they were the “heroes” and you were the “heals” for trying to take something away from the community. Rather than defining you as the technological leader, you start a cat fight in a highly competitive market. However, developing new technology that is released first as a “premier business edition” for a fee and then allowing those features to make their way into the free and open solution has been further evolved and proven to be a valid model for free. RedHat’s Advanced Server was an attempt to be trying at least some aspects of this concept combined with the certification platform to be discussed later. SugarCRM is another success story relating to this model. They have a commercial version of their open source effort. The Sugar CRM “premier edition” seems to be working. At the time we were involved in Linux, there were very few enterprise customers. To bring the company the most value in recognition and market share, it was better to innovate on the technology and just release the code openly rather than use it to garner more revenue for an extended time. Today, things are different. Combing richer features with a subscription model for business works and is becoming more accepted. Technology innovation can result in a very positive revenue stream while allowing the generic product to be fully open. When we experimented with this model, using a pure technology innovation model was very difficult because a single company could not keep up with the innovation of an industry. At the time, the innovative idea did not bring the company a lot of revenue or even recognition; it just showed the competitors what they need to do in their next release. What we hoped to ultimately do was to open source the Unix kernel, create a complete Linux application environment around either the Unix or Linux kernel, and a build and development environment that would allow our Replicated Site customers to select which ever kernel met their application needs. Our intent was to merge this with the UnitedLinux base to enable a unique platform for business that would provide, ISVs, VARs and corporate developers with the single most scalable operating system platform on the Intel hardware. We knew this would still not be enough of to get us out of this dilemma was to specialize and truly focus on delivering a total value proposition to a customer in a given market segment. So we would then offer a premier package with additional richer utilities and services for our VARs and Replicated Site customers that would facilitate the installation and management of their unique services on top of Linux.
Collaborative Cost Sharing and Specialization
I include the concept of a collaborative cost sharing in this section, not because it is a standalone business model, but I honestly believe that the UnitedLinux collaborative cost sharing would have saved enough resources to facilitate the needed specialization to sustain a competitive business model. Clearly, by not spending time and energy duplicating our efforts on the same basic bits and bites, each partner in UnitedLinux could have focused on delivering real business solutions to specific customers that will drive further financial success. The other benefit of the UnitedLinux model is the ability to create a much better product then each of participating companies could individually. The down fall of this model was demonstrated by the actions of SCO. The collective success is also based on all participating parties continuing to support the founding principles. When one of the major players pulls out, for what ever reason, it weakens the collective whole. For this reason, many will choose to go it alone. Unfortunately, to remain competitive in today’s world will require interdependencies that most are uncomfortable with. Those who continue to experiment with and take the risks necessary to perfect this model will eventually win, though it may require some failures before it is mastered.
How then does one really compete and maintain a defensible position against competition? The answer lies in truly understanding the customer and all the activities needed to meet the target customers’ needs. Linking those activities together in a unique way becomes the key. This is not new. Other commodity based industries, like the airlines, have struggled with these problems for years. SouthWest Airlines has done a masterful job of identifying their customer and building unique linkages between their activities to make if very hard to compete with them in their segment of the market. Michael J. Lanning in his book Delivering Profitable Value refers to this as delivering a complete customer experience or set of experiences. He correctly explains that deciding what value proposition the company is going to deliver to its chosen customer is what should constitute the business of the company. SouthWest is a great example of this principle. They specifically chose the market segment they were after and understood completely what decisions would enable them to deliver the decided value or customer experience to that market segment. This discipline has enabled them to remain profitable in spite of major tragedies like September 11.
Delta Airlines Offers:
- “Major” airports only
- First-class seating
- Baggage transfer
- Varied fleet of aircraft
Southwest Airlines Offers:
- Avoids Major airports
- No first-class seating
- No baggage transfer
- No meals
- Entire fleet of only 737 aircraft
Bottom Line: Bottom Line:
Higher Value = Higher Pricing Lower Value = Lower Pricing
The most often requested change SouthWest receives is to have assigned seating and food. They choose to ignore those requests because they greatly increases their costs and thus force them to raise their price and jeopardize their on time departures which are primary reasons people fly SouthWest. There is a significant difference between being “customer compelled” or “internally driven” and in understanding the complete experience the company intends to deliver to customer and still make money. Lanning does an excellent job explaining how to avoid so many pitfalls associated with trying to deliver a true customer value proposition. A “customer compelled” company does not have the vision or leadership that SouthWest has to say no to it customers who are asking for food service on the flight. An Internally driven company thinks it understands what a customer wants because after all it talks to them all the time and their product managers have been in the business for years but nearly 80% of the time they are wrong. Having said that, I would like to list several key pitfalls I have watched companies, including Caldera, fall into:
1. Companies want to imitate everything about their competitors — assuming rivals know something we don’t or they want to have an expert like an analyst tell them what they are doing is correct. Unfortunately, this is more common than many may at first admit. This behavior occurs when management lacks direction or the leadership necessary to make the hard choices. Companies get so caught up in what their competitors are doing that they ignore their business, the customer.
2. Companies substitute operational effectiveness for strategy because it is more concrete and actionable. In other words, companies do what is easy, they continue to look at ways to cut cost out of the system rather then take the necessary gamble to develop new products that deliver real value to new customers. They continue to milk the cash cow even at the expense of new product development. Frankly, I experienced these phenomena at Novell. There was a core group of engineers that felt all they needed to do was make NetWare faster and more efficient. Unfortunately, Novell was unable to put the resources on new opportunities. These companies will fall victim to the Innovator’s Dilemma described by Clayton Christensen in his books and writings.
3. Companies become technology driven rather than having a “customer focus.” This is an example of what Michael Lanning calls “Internally Driven.” My personal experience is that some companies get caught up on delivering the latest and greatest technology. The stuff that excites the engineers but the real paying customer does not see its value. This is so easy to do in the information technology space.
4. Companies who are still technology driven but want to give lip service to “customer focus” find a customer who will tell them what they want to hear. This group is obviously a variation on the number 3. What happens after delivering a technology searching for a solution, they begin to say the answer lies in delivering value to a customer. The pitfall is that the technologies in the company run out and find a customer who agrees with their approach. Rather than taking the time to really understand the customer’s need, they seek for a customer that agrees with their approach.
5. Companies mistake “customer focus” to mean they must listen to just what the customer says he or she needs only. This is logical step after steps 3 and 4 have failed. Again, this approach requires little or no risk or leadership on the part of the company. They simply go out and listen to a customer and try to deliver a product based solely on that input. The challenge lies in the fact that the market and technology is changing rapidly. Many customers do not have insight into the new developments and how they would or would not benefit their business. They are busy running their own enterprises. I recall one CEO stating that if he developed exactly what is customer wanted, he would simply deliver a solution that is what they wanted 6 months or a year ago. This solution is the “customer compelled” company described by Lanning.
6. Companies often make no choice to avoid risking blame for making a bad choice often the result of having to choose between not doing one of two things. This is just another option some companies choose because they do not have the leadership to take a risk so they make a decision to not make a decision or they continue to focus on the things they know how to do as outlined in number 2 above.
To deliver a complete customer focused experience requires true strategic leadership. Those who choose this model must have the discipline to say “no” even to a customer when it will destroy the primary reason the customer is buying your product initially.
The challenge of developing or reestablishing a customer focused experience is often primarily an organizational one and depends on leadership
In many companies leadership has degenerated into orchestrating operational improvements and making deals as previously mentioned
General management’s core should be delivering a customer focused solution:
– Defining and communicating the company’s unique position
– Making trade-offs, and
– Forging fit among activities
Managers at lower levels lack the perspective and the confidence to maintain a strategy resulting in constant pressures to compromise, relax trade-offs, and emulate rivals
The executives’ jobs are to teach others in the organization about strategy — and to say no
Strategy requires constant discipline and clear communication
One of the most important functions of explicit, communicated strategy is to guide employees in making choices that arise because of trade-offs in the individual activities and day-to-day decisions
The solution to this
· A day in the life of a customer
· Brainstorm and determine what questions must be answered to reach the target audience
· Create a minimal viable product or a product that can effectively test and answer your critical questions or premises
· Test and monitor the audience
· Analyze the data and determine new questions that must be tested, alter the minimal viable product and repeat
· Once there is success, evaluate options (What is available?) on how to effectively deliver the solution
Build, partner, buy or integrate
· Clearly communicate to the company the results of the tests and the rationale behind the tradeoffs so they can make the choices
Much of the process described above to develop a truly customer focused solution has been well documented by Eric Ries in his book “The Lean Startup.” However, the forces that first exposed the need for continuous innovation where first encountered in the open source world but now apply to nearly all markets because of hyper competition. The advantage to this methodology in a highly competitive environment is that as the selected customer experience is delivered and the company is able to determine the needed tradeoffs or strategy to deliver that solution, their competitors cannot imitate them without having to make the same difficult choices. Michael Dell from Dell Computers after giving a presentation on his very innovative process for building and delivering basically built to order systems to a broad industry gathering was asked if he was concerned about his competition imitating his actions. He responded that he had absolutely no concern about his competitors imitating him. By the time they could duplicated the system, the industry has moved on and it would require them to alienate their existing customers and their value propositions. Delivering a complete customer experience to a targeted segment requires a company to specialize. Other companies must make the same tradeoffs in order to effectively compete. However, being able to obtain a real understanding of the customer’s desired experience and then make those internal trade offs requires real leadership at all levels of management, ability for the company as a whole to take risk which are disciplines most companies will need to acquire. Linux can play an invaluable role in delivering a complete customer focused experience because of its flexibility. Caldera was working toward this objective when I left. Books like “The Long Tail” are validating the fact that the economy is moving from hits to niches. Delivering customized solutions to specific customers who need those solutions. The traditional “barriers to entry,” old models to ensure success are being replaced by strategy and evolution. Determining and targeting the customer the company is uniquely capable and interested in servicing (not being all things to all people), assessing their needs and the best way of servicing those needs, and then get in the market and evolve as needed. Evolution does not mean the company must listen to everything the customer says but rather truly understand what the customer needs and adapt. I believe the companies that apply this concept will be successful even in this highly competitive market.
Certified Platform and Subscription Service
For any operating system to be deployed extensively by a reseller channel to businesses or into the enterprise, it must have two very core things:
1) Broad hardware support
2) Broad application support
Broad Hardware Support
Every hardware component that makes up a computer system has a piece of software that enables its use by the system called a driver. A standard PC or server is made up of many other hardware components e.g. a hard disk, a CDROM, a monitor, etc. Not only does a driver have to exist for each of these components, but they must be tested and certified to work together in a particular configuration. Computer manufacturers must test and certify that their new computer systems function with all the drivers and possibly operating systems supported. Computer manufacturers release new systems every nine to 18 months. Supporting an operating system on all these new configurations is very expensive because of the required testing and certification. Manufactures have a very strong interest in not supporting multiple operating systems. They must be motivated by either very strong market demand or pretty secure new market expansion.
This testing and certification is particularly true if the operating system is focused on a particular market niche that requires specialized hardware components. The software drivers that enable use of the latest hardware systems and components must be available shortly after the system or component is made available on the market or the market will not materialize. The hardware market turns quickly because of ongoing rapid innovation.
Open source does facilitate the development of drivers because the source code to the other similar drivers is readily available and can be utilized to develop the new driver. However, open source does little to reduce the cost of testing and certification. In fact, some could argue that it may actually increase the cost because the opportunity for change is introduced by having the source code readily available and rapidly changing. A concern raised by some software providers was if they certified to a single product and the source code were freely distributed, it could introduce additional support obligations and liabilities they were not interested in incurring. The business customer also does not want to have a widely distributed source base.
Broad hardware support refers to not only having drivers available for new hardware but having companies who have tested and certified those drivers to work well with their computer systems or products. Without broad hardware support, many application developers cannot even begin to write or port (move) their applications to a new operating system.
Broad Application Support
Applications are produced by software companies, value added resellers servicing a unique market niche or by the enterprise for in house consumption. Each new computer system and/or hardware device must also be tested to work with their existing or new application. Most software companies design their software to run across multiple operating systems and hardware platforms. However, each operating system and hardware device must be tested and certified to work with any given configuration. Value added resellers and some enterprises will significantly limit their choices of operating systems and hardware configurations to reduce their costs of this required testing and the complexities of supporting the different choices. Like the computer manufactures, these entities must be compelled to embrace a new operating system by either broad market demand or perceived, but quire real market opportunity. The challenge is quite difficult because it is cyclical. You must have large market demand for the hardware developers to write drivers and for computer manufactures to be willing to test and certify. Without the applications, there is very little market demand. The applications need the new hardware to take advantage of the new features.
Since the market dictates that Microsoft operating systems be supported on nearly all Intel based computer systems, testing, certifying and supporting another operating system on the same architecture is costly and unwanted unless it can bring new market growth. Eroding existing market share on the same platform is not interesting or compelling because of these costs. For computer, hardware and software manufactures selling the same number of units while increasing their testing, certification and support costs is not at all an incentive. Linux was able to gain acceptance because it grew with introduction of an entirely new market niche, the Internet. However, different distributions of Linux function differently when it comes to both hardware and software drivers and applications. The computer manufactures, value added resellers, application developers and enterprise customers will not tolerate more than two major flavors of Linux. Many want more than one because they do not want to experience the pain of working with another Microsoft, but they do not want more than two. Even with two, they now have to test and certify three different operating systems on the Intel architecture. Being one of those two platforms means that you can charge more for the “distribution” because it is tested and certified to work with all the major hardware and software solutions and the market cannot afford a third competitor on the same operating system platform. Both RedHat and SuSE Linux are utilizing this approach to charge corporate customers for their distributions through the subscription model described below.
Once a product has broad hardware and software support, an annual subscription service not unlike annual maintenance and support can be generated. Business customers are willing to pay an ongoing service fee to ensure that the “certified platform” is updated with the latest security fixes and software enhancements. The subscription services would ensure that the changes and enhancements do not cause existing applications or hardware drivers to malfunction and protect against security breaches. Caldera intended to include the Unix Kernel in the UnitedLinux offering so that while it would be open sourced and basically be a Linux application environment, it would provide developers, VARs and enterprise businesses with a uniquely scalable, certified hardware and software platform.
Several trends are occurring in the software industry that will support this subscription model. The traditional software cycle is to produce and release a new version every 18 months to two years. New versions are often full of bugs and have many features that most customers are not interested in. New versions require significant costs in training and require hardware upgrades to support the increase footprint. This cycle has fueled the software and commodity computer industry for years but many customers do not need this service for some of their server applications and will be free to load other non-certified versions of Linux. The movement to Software as a Service or SaaS, will also legitimize the concept of an annual subscription as your computing infrastructure is in the cloud and not on your physical premises.
Companies that deliver digital products will have to find a way to make money when their core service is offered for free just like Linux. However, these same lessons still apply for nearly all other businesses that will have to find a way to compete in an increasingly competitive market where margins will shrink and their need to make a profit on decreasing product life cycles. The principles gleaned from Linux can provide guidance on how to differentiate, specialize and continue to provide real value to their customers.