Open Source Commentary from Navica's CEO, Bernard Golden

If Open Source is Free Why Has My Job Gotten Harder? IT in the Post-Proprietary Era

In this issue:

The Impact of Open Source On Proprietary Software Vendors

Why Clayton Christensen is Wrong About Open Source Commoditization

So, if Software is Free, Why Has My Life Gotten More Complicated?

Takeaways for the Post-Proprietary Era

The Impact of Open Source On Proprietary Software Vendors

Last month's newsletter discussed the impact of open source on the proprietary software industry. Briefly summarized, it forecast three outcomes as proprietary software companies confront open source:

Reduced margins

Proprietary software company margins will shrink as free alternatives become available. When your competition makes an equivalent product available at a lower price, generally you have to drop your prices. As free open source products become significant competitors in software market segments, incumbents will need to respond by lowering what they charge for their products. Therefore, those juicy margins historically associated with proprietary software will trend downward.

Unbundled services

Software companies have traditionally included many services as part of a license purchase. Technical consulting is typically provided on both a pre- and post-sales basis. This takes the form of architecture reviews, roadmap discussions, prototype building, and so forth. It's easy to recognize these services as an add-on to the license purchase.

There are other deliverables not so easily identified as add-ons. Many times these deliverables are characterized as marketing. Whitepapers, stories placed in technology publications via public relations techniques, analyst briefings that make a company's product functionality comprehensible to end users all fall into this category. It's easy to overlook the utility of these deliverables and not recognize them as services.

Finally, of course, one service often unrecognized – indeed, widely disliked – is a vendor's sales force itself. Again, it's easy to dismiss the importance of this service, given the irritation factor of many sales people. On the other hand, a sales rep will respond to questions, lobby for other (valuable) services from organizations within the vendor, and so on. In other words, sales reps help IT organizations in the decision-making process.

Abandonment of the infrastructure

Last month's newsletter didn't put it like this; instead, it forecast that margins (and therefore vendors) would move to the edge – the edge of the software stack, edge of the organization, and edge of the culture. Seen from a different perspective, this implies a move away from the core of the infrastructure. Naturally, it won't be immediate, and the vendors will resist with all their might, but eventually, core infrastructure will be low-margin territory abandoned by commercial vendors and left for open source. The vendors won't have a choice – the capital markets will force them to. Infrastructure will enter a post-proprietary era.

Last month's newsletter pithily summed up the situation in its title – Enterprise Software: Requiem for a Business Model.

Why Clayton Christensen is Wrong About Open Source Commoditization

Many people describe the open source phenomenon as the “commoditization of software.” Indeed, discussions of this commoditization have become a cliché of the open source movement. Free software, it is proclaimed, makes this trend inevitable.

But, does open source truly represent commoditization?

A very astute observer of the rise and fall of products is Clayton Christensen. He has described the process of commoditization very perceptively in his books on innovation.

According to Christensen, commoditization follows a predictable pattern. Early in the life of a product, it is not very capable, so getting acceptable performance from it requires deep technical knowledge and a single provider to integrate all its piece parts. Naturally, this knowledge is not widely available, so providers are able to achieve proprietary margins for their expertise.

As time goes on, however, the product becomes more powerful and less technical knowledge is required to extract acceptable performance from it. It becomes, in Christensen's rather infelicitous phrase, “modular and conformable.” For most people in the software industry, a more common phrase would be “standardized.”

Once the product becomes “modular and conformable” little expertise is required to build the product. Margins on the product fall, and the interesting action moves to elsewhere in the product chain. An example of this process is the PC. When first constructed by IBM, lots of expertise was required to integrate all the different components. Over time, the components became more powerful and the integration mechanism (motherboard) became commoditized, so that the margins on the box fell. Today, component and motherboard manufacturers churn out products, but make minuscule margins. However, Dell has been able to take advantage of this “modular and conformable” platform and innovate via the direct build-to-order business model and achieve very healthy margins.

In summary, commoditization is associated with two distinct characteristics: (1) ease of use via standardization (being “modular and conformable”); and (2) low margins.

To repeat the question, does open source commoditize software?

According to the price criteria, open source certainly qualifies as a commodity. Free is about as low a margin product as can be imagined.

When we turn to the “modular and conformable” criteria, open source's commodity status is not nearly as clear. Software in general is very difficult to use, and open source may be even more difficult in that it typically lacks ease-of-use features like wizards, graphical administrative interfaces, and so on. Put more bluntly, it is an enormous challenge to install, configure, program and administer all software, especially in complex, integrated infrastructures. Software is nowhere near “modular and conformable.”

In other words, Christensen's theory does not apply to open source. It has fulfilled only one half of the commoditization equation.

A question remains, therefore, how is it that, despite not being “modular and conformable” software has dropped in price to free via open source? After all, Christensen's theory holds that if a product is difficult to build and integrate, vendors should be able to extract high margins.

The answer lies in the marriage of the fundamental nature of software and the rich technology environment that sprung into being during the past five years.

Software itself is a unique good. It is nothing more, really, than captured thought. A program is designed and created by imagination, and then is made available in a digital format. Software is a capital good with virtually no cost of goods. All software requires for manufacture is one or more smart persons who have learned a computer language and have the willingness to ... think.

The technology environment we live in today is astounding. It's easy to overlook how much power and connectivity is available to billions of people at incredibly low prices. A decade ago, a PC cost $4000; today it costs you about $500 for a machine easily powerful enough to support heavy-duty software development. Likewise, broadband connectivity is now widespread throughout much of the world.

What all of this means is that the ability to develop and distribute software, formerly limited by high hardware prices and poor connectivity, is now widely dispersed to anyone with the desire to do so.

To offer one example of this phenomenon, consider Teodor Danciu. A resident of Romania, he worked as a software engineer for a local company. Frustrated by the lack of an inexpensive Java-based reporting tool, he decided to write one himself. Working in his spare time, he began distributing his open source product, Jasper Reports, via the Internet. Four years on, over 700,000 copies of Jasper Reports have been downloaded, and Danciu himself (still located in Bucharest) now works full time on Jasper, having been hired by a venture-backed company commercializing his product.

Ten years ago, Danciu would have been consigned to frustration about Java reporting. Thanks to cheap hardware and broadband connectivity, his product now threatens incumbent report vendors. This is what is gutting pricing for software. And it's not going away.

So, if Software is Free, Why is My Life More Complicated?

Open source should be an unparalleled joy for IT shops. A critical input for their work, software, is dropping in price. Furthermore, they're no longer under the thumb of proprietary vendors. Thanks to open source software, IT shops can now obtain – at no cost -- as much software as they want, use it as they want, and modify it to better suit their unique situation.

A perfect world?

The fly in the ointment is the imperfect commoditization of software. The drastically reduced price of software – one half of the Christensen's commoditization formula – is certainly present. But the other half of the formula – the “modular and conformable” capability – is still missing. The margin necessary to support a range of services has disappeared, but left behind the arduous work of making the infrastructure work.

There is an impedance mismatch between the future unbundled software offerings of an open source-impacted industry and the still-needed integrated offerings of an infrastructure-challenged IT user base.

IT organizations are going to face an enormous challenge in coming to terms with the post-proprietary era. Today, we see only glimpses of it, but it's coming.

To illustrate how different the new software world will be, consider one very well-known open source application vendor. When it receives an RFP – the staple of IT decision-making – it refuses to respond. Its position is that lack of license fees precludes this kind of sales activity.

Instead, it asks the RFP originator “how much is it worth for you to get this information?” In other words, selling is now a distinct service, payable in advance. The IT organization has a choice between purchasing an external service to help it make the decision, or doing its own due diligence. The vendor isn't going to help for free.

In the future, this kind of scenario will be played out time and again throughout the IT value chain. A whole set of IT operating assumptions will be vaporized by the impact of open source on software pricing. It is no exaggeration to say that open source will cause more change in the software industry than any previous development.

Takeaways for the Post-Proprietary Era

Open source is causing a sea change in the IT industry. As margins fall, the nature of vendor/user relationships will be transformed. However, just because software is cheaper doesn't mean it's any easier to use. Therefore, it's critical that IT organizations get ready to confront the post-proprietary era.

The breakdown of today's economic assumptions and relationships will challenge IT organizations and force them to reconstruct their assumptions about product selection, project management, and expected service providers.

Next Month: Six Critical IT Skills for the Post-Proprietary Era