On his own blog, Schwartz says he doesn’t buy the argument that the computing industry is commoditizing. But as host of a panel at the AlwaysOn 2004 Innovation Summit, and later in a conversation with ZDNet contributing editor Steve Gillmor, Schwartz identifies bandwidth as the commodity Sun is targeting.
ZDNet: How do you use commoditization strategically, and how is it used against you?
Schwartz: In general in the industry, you ascribe the word “commodity” to your competitors’ problems, and that enables you to engage in a theatrical debate about whether or not their pricing is appropriate, or whether or not they’re overcharging, or whether or not they’re locking you in. But I thought it was telling that Rod [Smith] from IBM is willing to admit there’s no binary compatibility across Linux and that’s a bit of a fantasy; that’s not the layer at which you expect to find interoperability.
So you hurl the word “commodity” at somebody you’re trying to disrupt, and that’s becoming more evident and transparent by the day. And as Sun delivers more and more competitive products at more and more disruptive pricing models, to the extent that the market does in fact commoditize—and I think right now the only commodity that I would agree in any way, shape, or form to in fact be a commodity is bandwidth—then those that are in a position to really harness intellectual property business models, [or] strong balance sheets, are going to be the ones that best leverage that word.
It’s been thrown at Sun, and it’s been kind of hurtful for the last couple three years--but as we prove to be the low-cost supplier of middleware for the Java Enterprise System, with the desktop for the Java Desktop, with x86 servers, with the evolution of our storage devices--a lot of this will inure to our benefit, and we can go back on offense, which is a much more comfortable position than somehow being derided by the word "commodity."
Yet, you pointed out that none of the panelists were in a position to take advantage of bandwidth…
Of that specific commodity, and therefore what they would necessarily need to focus on is in fact what I think we’re doing, which is competing. And we’re competing on price, technology, license models, moral values, social utility. A lot of the rhetoric in the open source movement is really to describe a social movement called free software.
When you look at the price Red Hat is extracting from the market, their recent tumble notwithstanding, or you look at the comments that the individual in the audience [during the panel] made about users feeling locked in to MySQL, it’s partially a degree of naïveté in the sense that, yes the open source movement totally radically changes the way we should all think about intellectual property. But, some of the fundamental dynamics of the industry aren’t going to change. Businesses are still going to qualify to a stack of software; there will still be value and utility in the purchase of that software so you don’t have to employ 500 people like some online retailers are doing to build operating systems.
Is that outsourcing at a platform level?
Well, I think Google is the only company I know of that builds their own PCs. And I would quibble about whether they achieve a remarkable competitive advantage on that fact.
You did quibble about that [in the AO2004 panel].
And I will continue to do so [laughs]. There’s a reason why they [Google] run their billing system on a big Sun server. Why? Because they didn’t want to build the PCs to do that. They bought MySQL. Why did they buy MySQL? I’m sure they’ve got all those PhDs; they could have written their own database. The point is Google is like every other business out there: they’re going to make investments.
It’s a combination of build and buy.
Yeah, and everyone always will be, unless in fact their businesses, as [IT Doesn’t Matter author] Nick Carr might assert, are so uniformly undifferentiated.
You’re saying there’s no fundamental difference between the three search players?
Today there is no fundamental difference. But look at eBay. What’s the fundamental difference between eBay and MSN Auction? Well, eBay has more people in it. Let’s just focus on that. That’s a differentiator—it’s a huge differentiator. Why did they get there? Because they had a different approach and they were first to market.
So does that mean that once auction is a commodity according to Nick Carr, eBay should throw out their auction engineers? That’d be dumb, because those engineers are probably responsible for getting the next million customers. The appetite in the market is to say the IT industry is commoditizing, therefore don’t pay attention. Which is dumb, because even though there are some components now which are very readily acquirable—storage, basic computing, basic operating systems—there’s still a huge amount of value in them. Huge, because it’s become universally demanded. Has it become a commodity? No, I think that’s a bit of a farce.
In a previous panel at AO2004 on search, each player was promoting their little piece of the pie and how they were leveraging it. AOL has this incredible amount of events-based data that they’ve captured, even though it’s shrinking. Is that viable or a misguided subset strategy that isn’t going to play against a federated model?
There’s enormous value in the federation of these networks. My two favorite examples: A phone that doesn’t roam on the GSM network, or even on a CDMA network, isn’t a very useful phone. An ATM card that only works in your bank’s ATM machine isn’t very useful. So federation in the long run always wins. The companies that fought over the standardization of railways or media standards—in the end, they always win.
The commoditization of identity—how does that happen?
That is, I would argue, the most interesting and the most contentious. The standards around identity may in fact emerge—where Project Liberty is heading is sharply in the right direction. But identity is strictly bound to a customer’s relationship to a business. What’s the difference between Visa, MasterCard, and American Express? Because they, by the way, are arguably the largest providers of identity services in the world.
The seminal conclusion that I’ve arrived at over the past 20 years in this industry is that exactly those trends that have affected the consumer industry--that brand, number one, and convenience, number two, are the determinants of success—will in fact now replicate themselves in the technology market and will matter enormously, as the promise of service levels or stability or security and convenience will matter enormously as well. Customers don’t want to invest a ton of time, whether it’s assembling middleware or federating their identity systems. They just want to make a purchase with a credit card.
Are we at that point?
Nowhere near it. Google has Gmail, AOL has AOL, Microsoft has MSN, Amex has Amex.
These strategies of "Oh, we have this wonderful walled garden that delivers this incredible amount of data" don’t have a lot to do with the other walled gardens sitting right next to them with different characteristics.
At the consumer level, federation is called multiple credit cards in your wallet. At the business level it’s the retailers who can accept MasterCard and American Express. Where are we in that model? It’s not going to be so clear. There’s a spectrum. One thing is becoming evident to me over time: consumers are increasingly in charge, not businesses. I don’t care what walled garden you build.
So we’re getting away from customers to consumers. The difference between a personal user and a professional one is non-existent, isn’t it?
There are a billion wireless devices sold last year, 350 million of them running Java, and the one thing we found after extensive study is the vast, vast majority of those people have jobs. You tell me the difference between a consumer and a professional.
It’s time-slicing.
Exactly. Do you want to carry two phones, though? No. You’re going to have one that serves both purposes. Look, if I listen to The Gillmor Gang-—am I doing it at home on a weekend because I’m at work or because I’m interested in whom you are interviewing? I don’t think the delineation between your work life and your home life is nearly as clear as it used to be. One of the artifacts of the pervasive network is you’re doing whatever you want to do whenever you want to do it.
In the opening conference keynote, FCC Chairman Michael Powell talked about finessing the transition from the entrenched carriers with their legacy spectrum investments to a hybrid that includes unlicensed spectrum. Will we get to a model where we string a WiFi network down 101 so that Gillmor Gang can be listened to in a car held captive in a traffic jam?
Whether that is the right approach, or just having the Gillmor Gang available through XM Radio or whether it’s going to be available through the GSM network because AT&T and Cingular put it up…there’s no one answer. I have a lot of respect for what Michael Powell is trying to do. I think for the most part he’s directionally correct, but to try to manage the evolution of that process in a heavy-handed way …. Look, Skype doesn’t pay attention to Michael Powell. And neither does the vast majority of the innovation that occurs on the network in the traditional software development perspective.
What are the propagation vehicles for that network? I think they’re as dynamic as the content on the network itself. I don’t see any slowdown. And whether it’s the evolution of EvDO (Evolution Data Only, the optimized version of CDMA 2000) networks or the propagation of what Wozniak’s up to, it’s not clear to me that anyone is really in the long run is going to be able to erect a walled garden that prevents the transportation from one network to another. It’s going to become more ubiquitous as time moves forward, and what will ultimately attract the majority of customers will, again, be brand and convenience.
