Snorkel is the clever name Miko Matsumura has given the proposed combination of Sun and Oracle. I blogged my first impressions about this deal when it was announced on April 20.
My biggest concern was, and still is, the culture clash between Oracle’s execs and sales force and Sun’s and its customers’ vision. After reading through the mostly vituperative discussions on the Oracle developer forums and the Java forums, and stimulating a spirited dialog among my own circle of IT architects: Patty’s Pioneers, I am even more concerned about this merger than before. For three reasons:
1. I actually love Sun’s vision (at least as articulated by their CEO, Jonathan Schwartz) of the future of computing. I believe it’s pretty accurate.
2. I believe that Sun is more in synch with its customers’ self image and goals than most companies are.
3. Although Sun and Oracle have many customers in common, Oracle has a much more adversarial relationship with its customers. This does not bode well.
Jonathan Schwartz’s Vision
The IBM deal was apparently Jonathan’s deal. The Oracle deal was presumably Scott McNealy’s deal. Jonathan did a good job of articulating and selling Sun’s vision in a series of blog posts on Sun's Strategy in March. In fact, these posts are so compelling, that they may have been part of what “sold” Sun to its suitors. What has always interested me about Jonathan Schwartz is that he is the “uber-geek” that most of Sun’s customers aspire to be. Making him CEO, rather than simply CTO, was a tremendous risk for Sun. He may not have done as good a job at running the company as he has done in articulating the company’s vision. But it’s a vision that rings true and which probably aligns with that of most of Sun’s customers.
In a nutshell, he seems to be saying: Our customers are developers and deployers of serious systems. We entice them with free, open software and tools they can use to design serious software. [We buy the precious attention of smart developers by providing them with great tools.] When they’re ready to deploy that software, they need serious, scalable, robust environments and support. [Sun charges for commercial quality software, service, and hardware, all of which you need when you deploy for real.] That’s where we make our money. Because we have lots of software tools that developers all over the world love (Java, MySql, OpenSolaris, Glassfish, NetBeans, ZFS, etc.), millions of developers all over the world download and use our tools. By seeing what they download, we can tell how serious they are. This gives us an amazing pipeline of leads that we actually don’t have the wherewithal to pursue. [Hint: if someone had a bigger sales force, they could convert many more of these leads].
We have hardware and software (O/S and middleware) that performs well and is well-loved and appreciated in the server market. We can use these standard servers and our open APIs to compete and win against proprietary players in the networking (Cisco) and storage (EMC) markets. This is a $160 billion market. [You need standard, not proprietary hardware, AND standard software.]
For the millions of downloaders of free software and tools who will not be building data centers, deploying enterprise software, or running supercomputers, we have SunCloud. They can deploy their software in the cloud and pay as they use the compute cycles. For our customers who have large data center deployments, they can also have virtual data centers in the cloud, with Sun’s great engineering to keep them high performing and secure.
That’s what Oracle thinks it is now buying. Unfortunately, many of the customers won’t come along for the ride.
Sun’s Customers’ Self-Image
Sun’s customers are IT architects, developers, and deployers of commercial and enterprise software. They are very smart, quite opinionated, and have an engineering mindset. They believe in simplicity, good design, performance, and robustness. They are strong proponents of open software. They like business models that make sense (are fair) and that don’t have a lot of friction (heavy negotiation) associated with them. Sun’s customers like Java, open source software, and a business model the lets them evaluate and use tools for free that save them time as they’re designing enterprise or commercial applications. When they’re ready to deploy that software, in a Data Center or in a cloud, they want high performance, robustness, and, again, open standards.
Oracle’s Image with Customers
Oracle’s software products are fine. As are the brands that Oracle has purchased (Peoplesoft, Siebel, et al). That’s not the problem. The problem is that the Java community and the open source community see Oracle as a nasty, proprietary, predatory player. They don’t like Oracle’s heavy-handed selling, and they don’t trust Oracle’s stewardship of open source components. They expect Oracle to jettison Sun’s hardware or to sell it off to HP.
Can Oracle Maintain the Java Customer Franchise?
You can buy the code, but you can’t buy the customer franchise. Java is a strong brand. Note that Sun’s stock symbol is JAVA. It’s not just the programming language that developers care about. It’s the spirit and the community and the practices that have grown up around Java. The deployment execs who trust Sun’s hardware also trust the robustness of the code their developers create. One CIO of a financial services firm summed up the negativity floating around in this way:
“SUN was a partner and had a lock on our enterprise servers. Oracle is an adversary with predatory pricing that you had to use. Will never buy another SUN box. Java strategy now in question.”
That pretty much says it all!
RSS 2.0 Feeds
Wow! Great post, Scott! (Scott is also one of my "pioneers"...I love the Apple analogy and wouldn't have gotten it, if you hadn't pointed it out.. I wonder if any of the "mainstream media" will figure this out??
Posted by: Patty Seybold | May 03, 2009 at 11:07 PM
An angle missing from most commentary on Oracle's Sun acquisition is to note the many fresh parallels of the combined company to the business-model of a certain other well-known enterprise. See if you can figure out who I'm talking about.
Consider: In a commodity market, this company has distinguished itself through vertical integration that includes both proprietary in-house and open-source software, some superb applications, innovation in the field of cloud resources, a pan-format OS architecture, extensive and risk-taking hardware engineering and manufacturing--even chip design with last year's still-mysterious acquisition of PA Semi--and arguably the finest design smarts in each niche it has chosen to address.
It's Apple, of course. The corresponding elements in the Sun/Oracle combination are obvious when you think about it. The main element that doesn't quite match up is on the sales front, where Apple has both third-party and online dealers and its own retail network (itself a differentiating reason-to-buy for its customers, as I've found recently when I needed help with my son's 4 1/2-year-old PowerBook and my own Macbook Pro. How refreshing to be able to talk face-to-face with someone possessing a clue). Meanwhile Oracle has its global network of sales professionals and consultants, including the Siebel organization.
The result in Apple's case is an artful melange where the whole is greater than the sum of the parts. In particular, in a field awash with commodity products and with its competitors' support costs ballooned by the obligation to support all that diversity, Apple's strategy has been to differentiate by squeezing-out the best possible performance and user experience while containing support costs (what I call "The Great Unfunded Liability") through control over a limited number of components and configurations. Intriguingly, for the broad class of commodity peripherals like printers, Apple seems to rely on the open-source community for drivers and other resources.
(There's one other business-model element that doesn't match up: Oracle now owns a leading virtualization player. What to do with that, Larry? ...How about making your next acquisition a BIOS company? Owning the preboot space would be a marvelous place to be for your newly re-vectored software/hardware company and would monetize what is currently a free/open-source endeavor.)
It is worth noting that Larry Ellison is a very close friend of Steve Jobs, and they've spent some quality time together in recent weeks during Jobs' recuperation. Perhaps Jobs gave encouragement for Ellison's grab of Sun. The resulting corporate combination certainly looks like something Jobs himself might have crafted.
One could point to IBM as another place to look for similar parallels, or HP, but there's a certain flashiness to this combination that separates it from those worthies. Frankly, the world of big-iron IT hardware and software could use a little flash and sizzle... and Apple-class innovation. Lack of software development tools' support for multicore architectures, for example, is a frequent lament among IT pros, despite nearly a quarter-century's example of how to do it provided by the personal-computer world in the form of LabVIEW. Which was, let's remember, first developed on and for the Mac.
Unlike some of my colleagues and friends, I'm bullish on the new Oracle.
Posted by: Scott Jordan | April 30, 2009 at 01:18 PM
From Peter Horne in Sydney--One of Patty's Pioneers:
I was quite surprised at the negative response to the Oracle takeover of Sun. I thought it was business as usual to some degree - the strong taking over the weak.
However the take-over does raise some far reaching questions for the commercial/open source debate. Has Oracle through this acquisition, debunked the notion that commercial Co's could be both commercial and altruistic in terms of their support of open source? What Oracle has done is snatch a fully integrated stack, much of it open source, back in to the aggressive take no prisoners commercial sphere. Those that were relying on these tools to stay in friendly commercial hands have had a wake up call and are now worried about a risk that was always there.
I think that "true", hard nosed commercialism, and open source are both winners. A strong commercial Co, has proven its strength over a weaker one, and open source may be re-invigorated with a new raison detre. The true open source advocates, actually the FSF, never trusted the commercial side anyway and will feel vindicated. Those who were kidding themselves in the middle - either that open source was commercial, or commercial Cos could be big in open source, are vulnerable in the middle ground. That includes IBM.
Perhaps this deal demonstrates the boundaries. Open source develops components, the commercial world integrates. The reason for this is that open source can't integrate as you need a business problem to define the integration problem, commercial software Cos can't compete as single problem specialists because once there is enough interest then a dominant open source alternative emerges. Sun tried to be a component specialist, some open source, and it didn't work. It will be interesting to see if Oracle can make the fully integrated model work.
Regardless, I think the deal is a defining moment and gives the IT world some intrigue and makes it interesting again.
Posted by: Patty Seybold | April 30, 2009 at 07:53 AM