Q&A: Jonathan Schwartz on Sun's open-source business strategy

In a wide-ranging interview with the Sun CEO, we dig into the company's open-source strategy and its belief that freedom today is revenue tomorrow.

Jonathan Schwartz Sun Microsystems

Jonathan Schwartz is a man on a mission. While at Linuxworld today, I took an hour to visit with Jonathan Schwartz, CEO of Sun Microsystems. After spending an hour prodding Jonathan with questions about Sun's history and future with open source, I was left with one clear impression:

Sun is rising, and open source is the driver behind its rebirth.

Jonathan is an executive who sincerely believes in open source as a fundamental business-model advantage, and not as a cheap complement to throw to the community in order to drive sales of "the real value." It's not a marketing gimmick with him. It's a strategy for winning. Jonathan, despite wearing a tie when we met, clearly understands the importance of community before commercial. Or, rather, he understands that community leads to commercial success.

As he stressed, the open-source battle is not between Red Hat and Sun. They are allies. Red Hat and Sun both want open source to succeed, and both want this phenomenon that started at the edge of the network to define the entire computing landscape.

And so I asked,

At what point did you think open source was a viable development and distribution strategy for your company? What was the trigger?

Early on at Sun I ran our developer tools business. There's no clearer place than that to see that if you don't drive adoption you won't get revenue.

But not from the developer. No self-respecting developer pays for software. The average start-up or corporate developer doesn't want to buy any software. Nor do they have to, because there are great open-source projects that do the same things that proprietary products do.

We distribute three to four million copies of OpenOffice every week, and probably have 100 million users worldwide. The bulk of these users are students, retirees, etc. I didn't have access to these through an enterprise direct sales force. The only way to reach them was through free distribution over the Net.

Most don't care about the license--they care about the money they're not spending. (Think about how bizarre Google's model would be if they charge 10 cents for every search.) But those students and free users of today are the corporate buyers of tomorrow.

Maybe, but isn't it easy for Sun to give away software, given that it's a hardware company?

We're not a hardware company. We're a systems company. Folks have a hard time with this because they want us to be something that we're not.

When I became COO I took out to dinner two of my senior members of the microelectronics (chip/semiconductor) team. I told them, "Look, I'm a software guy, not a hardware guy. I need you to teach me your business." To this they responded, "Don't worry about it, Jonathan. It's OK. It's all software."

People want to segment out the software from the hardware in our business. I don't think this makes sense. How much of Thumper is ZFS? Answer: It doesn't matter. The value is in the system.

We're a technology inventor. We build the technologies that we think are valuable and differentiating. We don't defy categorization because we think it's fun to do so, but rather because these are the markets we think we can succeed in. How would you classify Apple? [I stammered out a reply.] Exactly. You don't. You like your Mac, and the experience is more than just hardware or software. It's both.

OK, so it's nontrivial for Sun to release Java, Solaris, etc. as open source. But how do you determine whether doing so makes business sense?

The single most important question that you have to ask in the software industry is "Do you understand your industry's demographics?"

Please explain....

OK. We ship tens of millions of Java runtimes that run on billions of devices. But now I ask you, "Who uses Java?" My friend's child who plays Runescape uses Java. Google uses Java. Nokia puts it on their cell phones. My 82-year-old neighbor does for a heart-monitoring solution. Each uses Java, but each would describe their relationship with the technology in very different ways.

There are four fundamental questions/topics in open source:

  1. Open-source licenses and the availability of source code;
  2. The impact of free (as in cost) software;
  3. The value of brand. As Red Hat knows, Red Hat is indomitable because of its brand, not its source tree;
  4. Who's asking? The answer you give to an 8-year-old is different from the one you'd give to a CIO. This last topic provides the answer to the open-source revenue question.

Why? Think about this: In a year where Sun arguably moved more aggressively to give away more free software than any other company, we grew our software business by 13 percent. It was the fastest-growing business at Sun (and doesn't even include Solaris, which we don't yet break out). We pumped out more software last year than we have in the history of the company. We gave it away. And yet our software business grew by 13 percent.

How?

If you're a London developer that happens to work for a bank, maybe you have your laptop set up with Ubuntu. It's perfect for you, and it's free as in beer. But if you're the CIO at that same company, you're going to demand a support contract for Ubuntu (or Solaris) running on your mission-critical servers, because you don't want the risk of systems going down without backup.

I don't expect many college students, developers, or start-ups to spend a lot of money on intellectual property. I expect someone whose job is on the line if a system fails to spend considerably more than nothing. The key is figuring out the difference between one's market and one's community. They are not the same.

You made the shift to open source, but surely there was a lot of consternation internally over the shift. I can't imagine that the decision went down smoothly.

There were definitely people inside the company who questioned our approach. There is no foot-dragging on open source now, but still lots of questions.

When Scott (McNealy) and I decided four or five years ago that this was a direction we were going to head, that was it. It was no longer a democracy. Leading a company is about leading and making difficult decisions.

There was resistance, yes, and we've been respectful of the dialogue with the people who don't agree. There were serious questions that we had to answer. Keep in mind: this was wholesale inversion of a very carefully architected development process. Given how critical Solaris was to the infrastructure of critical safety systems worldwide (such a 911 systems), we had to help the developers understand that lives wouldn't be at stake in this move (literally). We had to convince them that we would improve, not degrade, security, performance, etc.

But the marching orders were clear: we had to change if we wanted to move forward. And so we have.

The result has been fantastic. We build the fabric of the Internet, and increasingly speak the language of the Internet. One customer recently told me that they love Sun because "we understand the Internet." We understand community, standards, etc. That's different from what we were hearing four years ago.

Sun seems to have a longer-range vision of its open-source contributions than others, which seem highly near-term tactical in nature. What do you hope to gain from contributions to communities like OpenOffice?

This is a serious question for us. Despite being a $14 billion company with $2 billion in research and development (R&D), every dollar is precious. We don't want to waste any money that could otherwise drive value for Sun.

Solaris installations globally Sun Microsystems

One of our biggest strategic challenges is that we have roughly 4,000 quota-bearing salespeople. To me, it feels like my biggest competitors have 4 trillion sales reps. I'm not going to beat them by hiring more sales reps. I'm going to beat them by leveraging the single-biggest sales channel the world has ever known: the Internet.

Think about this: nearly every Brazilian sees my logo. Every day. Not because a sales rep puts it in front of them, but because they see it on OpenOffice, which their government or they downloaded for free.

Customers now find us. When I see Sun's free software downloads in Western China or Siberia, two regions where we have no sales presence at all, and yet we're doing business there, I have to ask how my competitors are going to survive when no one knows about them.

[Jonathan then alluded to this blog entry and associated mash-up (above), wherein he notes:

There's no way we could match this kind of global growth by sending out compact disks or sales reps--free software allowed interested developers/customers to identify themselves to us--rather than the other way around. And allows us to build vibrant relationships and communities across the globe--based on a free basic update service for individuals or small businesses, and higher value offerings for larger businesses. We are meeting new customers because Solaris no longer requires our hardware (or a human being to deliver it).]

But how do you translate interest and volume into customers? Downloads are nice, but how do you get a return?

The question to ask is, "What portion of the world knows my brand?"

With the Java platform I'd guess we reach 20 to 30 percent of the Internet every day (powering the games kids play online, the intranet application at a bank, etc.). Each of these constituents may think about Java in different ways, but in each case Java "sells" my brand. If you believe that brand is central to the next wave of Internet monetization--and I believe it is absolutely central--then the more people that know my brand, the more benefit inures to me.

We distributed nine million licenses of Solaris in the last two years. I guarantee we wouldn't have been able to make nine million sales calls. Seventy percent of these licenses are on Dell, HP, and IBM. This gives us a great platform to build a partnership with these hardware companies, and it also gives us a great pool of users, some percentage of which will want to pay us.

Why? How do we monetize these? When that technology is run in a Fortune 100 company in a mission-critical app, the CIO will hunt me down to pay me money. The cost of downtime for them is huge compared to the cost.

If you could start over again from scratch, what would you do differently?

I would have done it in 1996, the day I arrived. It would have been much easier then.

Easier, maybe, but would the market have understood?

The market may not have understood, but it's not "the market" (immediate paying customers) we would have been after. It would have been the students and other "free" users that would have become the paying users of the future.

If you were to go ask the CIO of any of the large companies around here in San Francisco (The Gap, Wells Fargo, Charles Schwab, etc.) whether they worry more about their Internet-facing IT or their intranet-facing IT, they will tell you that they are much more focused on external-facing IT. The consumer Internet, as it were. And so they are increasingly concerned with areas with large consumer populations like China, India, etc. This is tomorrow's market opportunity. These customers may not pay today, but these "tomorrow markets" are won by free adoption today.

What is your advice for open-source CEOs?

I don't think Marten Mickos et al. need much advice. They're doing a fabulous job. The people who need the advice are the traditional proprietary software companies. They don't recognize the tidal wave that is encroaching on their shores. When you can go with Postgres and save $3 million to $4 million in building the next-generation consumer Internet service, there's clearly massive disruption in play. That disruption favors open source.

We're in the second wave of the Internet. The companies that will win will be those that define this next phase. Open source will define it.

What is the business model for this?

What's the business model? I don't know. But if you don't have adoption, it won't matter what business model you use. Companies that sell open source are prioritizing community and adoption over instant monetization. We will win.

So, back to the earlier question, slightly revised: what advice would you give to the CEOs of proprietary software companies?

  1. Revenue is a lagging indicator of developer adoption. If you have a lot of revenue today, good news! You did something right three years ago.

  2. Align your sales force's compensation to fit your model. When Sun made the shift to open source, our sales force was the group within Sun the most opposed to the change. We therefore changed our sales compensation model to compensate them on bookings, not revenue. But we also offered to compensate them on revenue. So, if they wanted to close a revenue contract instead of a booking contract, they got to choose between the bigger of the two numbers. It took a day for them to make the shift. Open source gives them a body of users of trial software that they never would have been able to reach before. Now they are the biggest promoters of open source at Sun.

  3. Understand the demographics of your market. You need adoption in order to get sales. You don't need to monetize every user of your software.

Any parting thoughts?

There are a lot of different ways that I could spend my time. It's a great feeling to be able to do right for the world and for my shareholders at the same time.

Fascinating, fascinating stuff. Sun is a contender again, and largely because of the vision and tenacity of Jonathan Schwartz.

Tags:
Tech Culture
About the author

    Matt Asay is chief operating officer at Canonical, the company behind the Ubuntu Linux operating system. Prior to Canonical, Matt was general manager of the Americas division and vice president of business development at Alfresco, an open-source applications company. Matt brings a decade of in-the-trenches open-source business and legal experience to The Open Road, with an emphasis on emerging open-source business strategies and opportunities. He is a member of the CNET Blog Network and is not an employee of CNET. You can follow Matt on Twitter @mjasay.

     

    Join the discussion

    Conversation powered by Livefyre

    Don't Miss
    Hot Products
    Trending on CNET

    HOT ON CNET

    Find Your Tech Type

    Take our tech personality quiz and enter for a chance to win* high-tech specs!