You are here

Tipping his hat to the open source model

Red Hat CEO Jim Whitehurst sees the movement creating the biggest change in society since the industrial revolution.

BT_20160423_ACREDHAT23_2238969.jpg
Open source works on the principle of crowdsourcing. You have tens of thousands of open source projects around Big Data ... But rather than saying what's the best of six, you can say what's the best among thousands?'

JIM Whitehurst's relaxed demeanour belies the fact that he earned his chops in the corporate world as a crisis manager who brought Delta Airlines out of bankruptcy in 2006, while also preventing a hostile takeover bid by US Airways. As the chief operating officer of Delta, he led the fight to save the iconic American airliner, which he did with aplomb.

Those crisis days are behind Mr Whitehurst, who is now president and CEO of Red Hat, the world's leading provider of open source, enterprise information technology (IT) solutions. He no longer has to think in terms of restructuring, layoffs and the next meeting with creditors. Instead, he dreams big and is excited by the fact that something pioneered by his company is on the verge of changing the world.

Mr Whitehurst joined Red Hat in 2008 and has led the company to become the first billion-dollar open source IT company, and at the same time moving the company's stock up three-fold since he joined it.

One may wonder about a move from a company that is very much embedded in what is seen as the old economy, to what was, in 2008 (and even now is), a very nerdy IT company with a nifty name, Red Hat.

Actually, it wasn't really a leap of faith for Mr Whitehurst, who has a computer science degree and is a self-confessed computer geek.

He recalls that when Delta emerged from bankruptcy, the creditors got to choose a new CEO to succeed Gerald Grinstein, who left in 2007. Contrary to expectations, the Delta board didn't pick Mr Whitehurst and he found himself looking for something new to do.

"I got calls from a lot of large companies that were in trouble because I had done a big turnaround. I remember talking to my wife, and telling her, 'God, this is depressing, there is this opportunity but it's losing money here and it needs restructuring there.' I laid off tens of thousands of people in Delta and it was dreadful and I didn't want to repeat the experience."

Around this time he got a random call from the then Red Hat CEO, who had read an article about Mr Whitehurst. The conversation, according to him, went something like this: "You probably haven't heard of this company, it's called Red Hat."

Mr Whitehurst replied: "Of course I know Red Hat; I have Fedora installed at home!" Fedora is an operating system (OS) based on the open source Linux kernel, developed by the community-supported Fedora Project and sponsored by Red Hat.

"They offered me the job and my wife told me then, 'this is the first time I see a real glimmer in your eyes'. The idea of creating something, to grow something, it was a different model than coming in to deliver something (in the case of Delta, taking it out of bankruptcy)."

It helps that Mr Whitehurst is a passionate believer of the open source movement, which promotes the use of open source licences for all types of software. In open source software, the original computer programming code is freely available so that users can modify it according to their needs.

"The whole model isn't about gouging IP (intellectual property) but adding value around it. It seemed (in 2008) like being a great way to be part of painting a new canvas.

"It was not about rearranging the deck chairs. It was a chance to go do something new and it far surpassed my wildest imaginations."

Laying out his vision of what is happening in not only the IT industry but the global economy in general, Mr Whitehurst makes a bold statement. He says that the model which has allowed the growth of big companies with tens of thousands - in some cases hundreds of thousands - of employees around the world, due to its competitive advantage, will wither away over the next 20-25 years.

To understand this point, we need to backtrack a bit and look at why Mr Whitehurst thinks that the Red Hat business model - where the company gives away its enterprise software for free but charges for customisation and maintenance - works so well.

According to him, Red Hat has "deconstructed" the traditional software value chain, made ubiquitous by companies such as Oracle, SAP and IBM, and reconstructed it around the fact that the software is free but customisation and support are not. "I think one of the misconceptions most people have about software is that the content (software) is the most important part of what a company buys. It's important but it's only a part of a much larger package."

Red Hat has introduced a free innovation model in which it gives away the piece of enterprise software and has, instead, constructed the rest of the value chain around the testing, upgrading, maintenance and other aspects that need to be in place for a company to consume the software, sometimes for as long as a decade, he says.

"I'm passionate about this for two reasons. One is related to the software and the other is related to the impact on our society and open information."

Mr Whitehurst notes that most innovation in software today is happening in open source.

"That is not because of the software companies but because of large IT users. That's one reason that Google, Amazon, Facebook, Twitter, LinkedIn and others do all their innovation on open source.

"Everything that's happening in Big Data is happening first in open source and that's because these large groups of IT users are contributing to that."

On how Red Hat taps into the trend, Mr Whitehurst says there's a need to find a system that helps enterprises to consume all this innovation because not many companies have the expertise to do so on their own.

"If we don't find a model to help other enterprises to consume the pathbreaking work being done by a few, the potential of all that innovation would be wasted and this is where Red Hat comes in."

Mr Whitehurst notes that large banks and insurance companies are starting to get involved in some of these open source projects.

"We are helping companies to do that. One of the dirty little secrets of Red Hat is that our R&D (research and development) budget as a percentage of revenue is higher than the average software company's. People say: 'How's that possible with open source?' Yes, the upstream development is done in open source but all the downstream engineering and the commitment to decade-long support for these open source software, all this requires a lot of engineering and a lot of work. Actually our R&D percentage of revenue looks like a lot of other companies'; it's just that it is not focused on upstream development but on sustained engineering."

But why is open source so good at innovation?

You get extraordinary technology from open source for two reasons, says Mr Whitehurst. "One is because users are doing it themselves and they understand problems much better.

"The other is the development model. If you think about the traditional development model, let's say there's a data analytics problem. You have Oracle trying something, IBM trying something else and a few other companies are also giving it a try. You may end up having six or seven different companies trying to come out with a solution. And out of that, one or two would win the market because they have the best solutions.

"On the other hand, open source works on the principle of crowdsourcing. You have tens of thousands of open source projects around Big Data. It's like throwing out seeds in an open field and only one in a thousand will begin to sprout and only one in a thousand of those will actually grow to be a big tree. But rather than saying what's the best of six, you can say what's the best among thousands?

"So ultimately that's an innovation model because you start off with so much more innovation and then users decide which one of these best meets their needs and then they build up from there."

Mr Whitehurst notes that mobile, social, cloud computing and Big Data analytics all have their origins in open source. "It's a self-reinforcing cycle."

Business models such as those of Google and Facebook couldn't exist without open source because the cost of paying for all that infrastructure, in a traditional business model, would never allow the companies to offer their services for free, he notes.

"And if you don't offer the services for free, that's a huge barrier to people doing things. So the fact that it's free creates increasing use which creates a requirement for something new - for example, in Google's case, a database that is larger than anything that exists as there's a need to index the whole World Wide Web.

"No vendor offers a database that could index the whole Web. So Google had to write it themselves. They called it Big Table. They published a paper and Yahoo! said, 'I have the same problem' and so they took that technology and created Hadoop." Hadoop is a free, Java-based programming framework that supports the processing of large data sets in a distributed computing environment.

Mr Whitehurst adds that these large IT users are running into problems and solving them even before the traditional vendors know those problems exist.

This innovation is having a massive catalyst effect, he adds.

"We are getting into changing the world scenarios. What we are seeing around human genomics wouldn't have been possible without the Big Data innovations that happened because Google needed to index the World Wide Web. The fact that it's open allows these technologies to hop from one use case to another."

Mr Whitehurst notes that stock exchanges and militaries of various countries are using open source not because it's cheap but because it offers higher performance.

"It's a little bit of the innovator's dilemma. Open source started because it was a low-cost alternative. But as soon as you get this mass of users who are also contributors, you hit this inflexion point around innovation. That is why you are seeing a lot of the traditional software companies facing a lot of trouble."

The Red Hat CEO feels open source is also creating a system that allows for a whole different layer of collaborations to take place. If there wasn't open source, companies such as Uber or Airbnb couldn't exist, he adds.

"Take Uber as an example. If you think about the company, 99 per cent of the people involved in that system don't work for Uber; they choose to participate because it works for their schedules, there are lots of moms who will drop their kids at school and then go on and work as Uber drivers till it's time to pick up their kids back from school."

It's fascinating when you think of this type of self-emergent system, he adds. "I used to run an airline; we had thousands of well-educated people running all these logistics programmes to optimise the work employees did. Uber does none of that. If you thought about Uber as a closed system where all of their drivers were employees, they would have had to have massive linear programmes and models that ask 'How do we optimise these people and where's demand'? Uber doesn't have to do any of that.

"They just say we offer this service as a system and people optimise their own lives around this system and decide when they work and how much. Price varies along with volume but this whole idea that you have this self-emergent system that can self-organise to solve the problem - it's a huge deal."

Airbnb is another example, says Mr Whitehurst. Airbnb, like Uber, creates a structure. "You do not have to think about maids and room occupancy and all that stuff. The system ultimately handles that."

Mr Whitehurst adds that we are seeing a new value chain emerging in industry - Uber and Airbnb are just two examples - that is a lot like the open source system where there are points where things come together.

"In Linux there is Linux kernel.org, controlled Linux Foundation but you have all upstream stuff and the downstream stuff. One could argue Uber and Airbnb are examples of that in their respective categories.

"So what are the boundaries of an organisation going forward in a world where information and content are open or can be made open? Institutions as we know them classically today will not exist in 20 years.

"In the case of large multinational corporations, you will see more atomisation. You can use information to orchestrate the value chain and the logic of having tens of thousands of people in one company will start to melt apart. You will still have companies where it makes sense to have a lot of employees getting together but I think there will be no need for these large mega companies anymore."

Mr Whitehurst points out that this creates its own set of issues.

Today's societies are a reaction to the power of mass manufacturing which was the first time the world saw a huge change in GDP (gross domestic product) and per capita income due to an increase in output driven by economies of scale, he adds. "So these large corporations were born because there was value in adding relatively unskilled labour with massive amounts of capital. If you think of geopolitics of the late-1800s and the 1900s, it was different countries' reactions on how you split the value these large corporations created.

"And so whether it was communism in some countries, where you say let's collectively share, or socialism in Europe, or the economic welfare state in the US, where you had minimum wage and healthcare benefits for workers . . . It was all about how do you extract some of those dollars out of the capital holder and make sure it was shared equitably among workers."

The Red Hat CEO adds: "Now you are saying workers will no longer be part of these organisations. You will have many more individual actors out there. Well how do you do things like healthcare, tax policy or social safety nets? All of these things will have to be re-thought in a world where fewer people work for large institutions."

In the case of the software industry, Mr Whitehurst asks: "Do you need an Oracle, IBM or even a Red Hat, or do you need small nimble players to help make open source software consumable? You can go industry by industry and say there are some things where scale is still important but they are becoming fewer and fewer."

Eight years on the job and Mr Whitehurst still finds it exhilarating. As he says: "I love it every day. I really do. At some point I may get tired but I can't imagine doing something else . . . It's really a phenomenal place.

"I look across industries and in technology and I can't imagine a better place to be. You look at the big companies, they are all kind of struggling, either doing layoffs or recasting their business models. And then you get the really small companies which can be interesting but their ability to have a big impact is more difficult.

"But Red Hat is in this interesting position - everybody knows who we are but we have very little legacy and so we are focused on the future . . . We understand where computing is going, which I think is towards a really differentiated model. I can't imagine a better place to be."

It's little wonder that Mr Whitehurst has a lot of staying power - he runs a lot. He hasn't done a marathon in two years but he runs 10-12 miles (16-19km) during the weekends and around five miles most days.

His other passion is to hack around with his computer and his latest project is to get his family - he and his wife have 14-year-old twins - to write an app for both the Android and Apple platforms.

So which platform will be the first? Mr Whitehurst ruefully says it's likely to be the iPhone app which will be first out of the post.

"I'm ambidextrous as I have both an Android and an Apple phone. The problem is that Apple's Xcode doesn't run on Fedora and so I use my wife's computer, which is a Mac, and my family loves the Mac."


JIM WHITEHURST

President and CEO, Red Hat

1967 : Born in Georgia, USA

1989: BA in Computer Science and Economics, Rice University

1994: MBA, Harvard Business School

1985-1989: Programmer, Georgia Power

1989-1992: Associate, The Boston Consulting Group

1994-2001: Consultant, Project Leader, Manager, Partner, Boston Consulting Group

2001-2002: Acting Treasurer, Delta Air Lines

2002-2004: SVP, Finance, Treasury & Business Development, Delta Air Lines

2004-2005: Chief Network and Planning Officer, Delta Air Lines

2005-2007: Led Delta Air Lines out of bankruptcy as Chief Operating Officer

Since 2008: President and CEO, Red Hat '