Impressionistic transcript of a session from Always On on the Commoditization of Software hosted by Jonathan Schwartz, President & COO, Sun Microsystems. Jonathan has a related post on his blog.
* Rod Smith, VP, Emerging Technologies, Software Group IBM
* Marten Mickos, CEO, MySQL
* Miguel de Icaza, CTO, Novell (didn't make it, but this is an interesting article)
* Roman Stanek, Founder & CEO, Systinet
* Zach Nelson, President & CEO, NetSuite
Roman: People confuse commoditization with standards. Standards and APIs create a famework for innovation, it helps us built something competitive with larger players. Software is commoditized and innovation happens.
Rod: Happens in different places in the software stack. For customers, they want the function with value on top if it. Building on top of open source for another value proposition. Customers get confused with interop and commodity. What functions do we build upon and provide greater value propositions?
Jon: Commodities are the products you don't ship and your competitors do.
Marten: We look at it from the customer perspective. They want something that is quality, standardized, so we don't have to spent our time on that component. Within the company, trying to make it so easy the customer can see it as a commodity -- usable.
Jon: Useful as a marketing utility, but not a recruiting utility?
Marten:Last few years, no problem with either. Making things simple is the commoditization process.
Zach: Ubiquity does not mean commoditization. We are also an ASP, so everything we do is not a commodity, managing software is a commodity. Building our application is not a commodity. IBM Global Services has the infrastructure to operate commodities. Applying applications is not a commodity, managing them is.
Rod: Its good for us to have some things commoditized, it opens things up and drives down the cost of intergration. Doesn't mean free, there are business models to build upon.
Jon: Commodities means a product or service for which there is a universal or perpetual demand. Double edge sword, marketing benefit, but loose focus on innovation.
Rod: Linux early on was a fairly large affinity with developers externally. Helped grow the market. You can look at utilities differently.
Jon: We moved to a pricing model on a per employee basis to subscribe to the market structure. The danger is that there are still differences between products. In open source, can you can move from one product to a next -- substitution.
Marten: Comparing to other industries, e.g. Beer -- some people are selective, some people drink any beer that's cold. Software industry will go through similar phases. For many users they don't care so long as its easy, others need differentiation.
Roman: How well the tech supports changing your business model is left to innovative products.
Marc asks about DIYIT
Zach: From a focus group of CEOs we asked what worried them: everyone knew their infrastructure, the n ightmare of integration, biggest problem was fragmentation of databases. Every CEO is a software CEO because of integration and complexity. We try to reduce that complexity.
Rod: Entering an age when IT doesn't decide what the user will use. Cost reduction benefits.
Jon: If a customer believes its a commodity they buy it. If rocket science they run away into the hills.
Marten: Customer standpoint, mother told me that's what I have to focus on. For them, innovation for the customer, going beyond ERP, focus on new applications. Explaining this internally is the seeing simplicity on the far side of complexity.
Nicolas Carr has either done us a huge favor or the opposite. Entering a mode where network and bandwidth is a commodity with wide deployment. Opportunity is managing the resulting complexity.
Zach: Web services will not commoditize applications. In Siebel's UAN, their diagram is so complex, its amazing customers believe it can work.
Rod: Easy to go to a vendor and ask if you can support XML, its lowering a barrier. Provide an overall service and a bigger value proposition. I can integrate common functions without risking my whole business.
Jon: ASCII is a commodity.
Q: Internal tension between interop and innovation?
Jon: Standards are a double edge soward, a new company can say they support J2EE and be credible, larger companies are tougher. They provide a platform for innovation.
Roman: IT moves in steps. People interoperate, then innovate, then tension increases and then a new level of interop. Next couple of years there will be another standard to bring vendors together.
Jon: Software is subject to network effects and tipping. Do the service models break apart the tipping? Things tip to the lowest common denominator. Moving to models that break the ability to lock in and enable more transparency and substitution.
Rod: Virtualization and other approaches have different interpretations by different customers. Tipping points start to happen and customers pull in new directions. 3-prong works in the US, but not abroad.
Zach: BI take extracts from Siebel and SAP, put it in a database and run reports. SarBox combines up to 50 databases. Multiple datastores in decline.
Jon: Sarbox is Y2K in perpetuity, the biggest boom for the industry. Now I have a great threat called a 404 authorization to drive spending. As network commodities, vulnerabilities are exposed and architecture starts to have new concepts like security from day one.
Rod: We treated security as a baseline requirement.
Jon: Marten looking to shrink the size of an industry or looking to expand it to other parts of the world.
Marten: You must be able to abandon something old. I do it arrogantly to evoke emotions for change. We are expanding the market. Tipping point when you have volume.
Roman: Google is interesting because there is such a better enabling technology that they can provide applications on top of.
Marten: They use commodity components to be more competitive.
Jon: They created their own disk drive.
Roman: The innovation is how they combine inputs
Rod: But the underlying commodities are opaque to users.
Jon: SWT and Eclipse is choice, competition, its good. Absent of choice, you have a problem. None of us believe markets are shrinking.
Rod: Continue to be innovation on the client side. Microsoft has created barriers
Jon: That's our partner you are talking about Rod.
Rod: We will have innovation and the heterogenerity on the client
Jon: Lock in is as misused a term as commodity. Locking in is qualifying to specific set of ingredients. We have concens that Java will tip the same way Linux will tip. Free software, but are they locked in? Great theater and rhetoric, a problem of our industry.
Rod: Innovation is good, how do we harvest that and make it something they can integrate.
Jon: MySQL is an open source database, but customers qualify for a specific release and specific patch.
Marten: You need stuardship of a product. Licensing and standards manage risk.
Jon: Software is subject to tipping effects and the focus is on the evolution of standards that enable competition and substitution. I want to ensure that people in innovation are not shuttered away from Commoditization as a term. You can make money and innovate.
I defined what commodity means and pointed out that we don't have cooperation on commercial standards, only technical standards.
Marten: There are commercial standards and cooperation, but only in open source licensing like GPL.
Right, that's a part of why there is so much innovation in open source, there is room for more commercial standards to foster innovation
Jon: As moderator, I get the last say: For software to be a commodity, intellectual property is a commodity, and I'm not buying that.