Jim Whitehurst, CEO of Linux vendor Red Hat, has never been shy about sharing his vision of the future of enterprise technology. But at a recent lunch in San Francisco, Whitehurst extended his outlook well beyond the next quarter or two - he looked ahead two decades to predict big changes in the balance of power in enterprise technology - and wonder about how we're going to get there.
Who Will Own Big Data?
In 20 years, Whitehurst predicted, Big Data will have gone mainstream. We will be increasingly able to use analytics to better orchestrate how a business collects, analyzes and distributes information. Bascially, everything important about how it operates.
Who will be the winners and losers in such a world, Whitehurst wondered?
Will there be a couple of big winners? Giant companies like IBM, for example, that are able to supply that kind of analysis to a wide variety of customers?
Or will we have lots of small, vertical applications that leverage insight into particular products and markets?
Who Will Own Enterprise Technology?
For the enterprise, Whitehurst said, the challenge is to use that Big Data to do everything from marketing to logistics and beyond. And then the question becomes who in the enterprise will be responsible for it? Will the Line of Businesss (LoB) owner get the data from something that resembles today's IT organization? Or will Chief Marketing Officers buy the technology they need directly from those Big Data service providers mentioned above?
If that's the case, Whitehurst asked, Why have a CMO and a CIO? Why not have a single marketing/analytics organization? Instead of having a separate IT organization, they merge, leaving just a rump IT organization that is primarily in charge of support.
Either way, Whitehurst predicted, it's the data scientists who are in charge. The only question is whether they work in a central IT (or IT/marketing) department or directly in the Lines of Business. Whether those essential functions are company wide or organized around vertical markets. "The lines blur," Whitehurst said.
Big Changes For The IT "Factory?"
How it plays out depends on which side does a better job of leveraging Big Data: "Will marketing get tech or will tech learn business?" Whitehurst asked.
The challenge for CIOs, if they want to remain in the picture, is to find a way to bring in business people - not just tech people, Whitehurst said. And to change the structure of their organization to deal with the new reality.
Fact is, Whitehurst said, most IT shops today feel like factories. They look like manufacturing operations, organizationally and culturally in terms of process and quality. "Marketing organizations don't work like that," Whitehurst noted.
Of course, marketing has its own challenges: "IT vendors try to hide the complexity," Whitehurst said, but "dealing with Big Data is still a different skill set from traditional marketing."
Information Changes Who Gets Rich
These aren't merely academic questions. In fact, Big Data is likely to become the core of many businesses, Whitehurst said.
He cited the CEO of a big retail chain who told him that, "my entire business is about 'How many red sweaters in size XXL should we have in a store in zip code  in March.' Everything else is seconday to that kind of interactive data query."
There's more. The creation of true information businesses leads to a massive bifurcation of wealth, Whitehurst said. As we build a new productivity system around information, more wealth goes to companies that produce the data-collection and analytics expertise - and less to the end customers, who end up being genericized. (Whitehurst credited Marc Andreessen for the idea that "Software Eats The World.")
Where will that expertise come from? Most companies will use outside sources, unless they can create a competitive advantage by doing it themselves.
And in that scenario, who is supplying the competitive advantage? Who makes the money when the real value is in the information analysis, not the end-user organization?
"In 20 years, I think it's decentralized," Whitehurst said. "Right now, XYZ industries may have a source of competitive advantage. But data analytics pulls value away from that."
Whitehurst came to Red Hat from Delta Airlines, and he uses the airline industry as an example. In the era of value management and global distribution, he said, airlines are losing the ability to gain competitive advantage. In an online booking system, "Your brand is reduced to a two-letter code at the moment of purchase." It's barely a factor in the buying decision, it's all about price.
That kind of wealth and power transfer will play out in industry after industry over the next 20 years, Whitehurst predicted. The share of GDP in software keeps growing very quickly, he said, and "very few companies can invest in IT that is better than what the leaders offer."
That means that the technology industry's share of global profit will grow, Whitehurst said, even as the industry becomes less structured, predictable and controlled. Instead, it becomes a more chaotic, wide-open marketplace, like Silicon Valley… or the open source community.
Not an entirely surprising viewpoint for the head of an open-source software company.
Bonus Question: Who does Red Hat Really Compete Against?
Forget Microsoft's ongoing battles against Linux. According to Whitehurst, "We rarely compete with Microsoft." The two companies address "different workloads" and address "different parts of the datacenter."
"We never compete on deals, Whitehurst said, "If they hate us, I'm not sure why." After all, "If it weren't for Linux, they would probably have massive Department of Justice issues."
So who does hate Red Hat?
Whitehurst fingers Oracle. "They're turned their guns on Linux," he said, "so we don't go to market much with them."
But Whitehurst claims that he doesn't think a lot about Red Hat's market share vs. competitors. Instead, he says, he looks at larger issues like Linux vs. UNIX, free vs. paid and how to address new enterprise IT workloads.
Lead image by Fredric Paul.