X
Business

Oracle's Sun purchase: Grading the results so far

Oracle's purchase of Sun Microsystems is a little more than a year old and the results thus far look decent overall. The problem: Quantifying Oracle's success with Sun since the acquisition closed Jan. 27, 2010 is tricky.
Written by Larry Dignan, Contributor

Oracle's purchase of Sun Microsystems is a little more than a year old and the results thus far look decent overall. The problem: Quantifying Oracle's success with Sun since the acquisition closed Jan. 27, 2010 is tricky.

Why is it so tricky? Oracle has melded software with hardware to sell integrated systems. The company has stopped selling many Sun boxes and nixed reseller agreements. Meanwhile, Sun has lost market share in the server market, according to IDC. Some of this share loss is by design, but it's hard to discern what share losses are actually good for Oracle. What's good market share loss versus bad?

Even Oracle's operating profit goal for Sun is a bit murky. For instance, Oracle trumpeted that it was "completely confident" that it will deliver $1.5 billion in operating profit from Sun by the end of the fiscal year at the end of May.

But on the conference call, co-president Safra Catz delivered a big asterisk to that $1.5 billion total. Catz said:

As I told you a couple of quarters ago, we are really no longer able to identify the exact contribution Sun has made to our operating profit this year, but based on our results we fully expect to exceed our goal of $1.5 billion in operating profit in the first full fiscal year. Remember, we paid $5.7 billion for Sun out of cash, and you know had we paid for Sun based on the [HP-3Par] multiple it would've cost us nearly $140 billion. Don't worry, we wouldn't do that.

Strip out that HP-3Par quip and you realize that Oracle's operating profit claim for Sun boils down to: Trust us.

Related: Oracle delivers strong third quarter, to hit Sun profit target

So how is Sun really doing? Simply put, we don't know. Historical data is moot. Oracle's hardware sales are increasingly wrapped into Exadata and Exadata machines. These engineered systems are hardware and software bundles. How exactly do you strip out Sun's results?

With all of those caveats in mind, here's a stab at grading Oracle's acquisition of Sun. Giving Sun a direction. Oracle gets an A on this front. You may disagree with what Oracle is doing with Sun, but the company has been upfront that the goal is to boost margins and cut any product that doesn't generate a profit. Sun hardware today is basically a nice optimized conduit to sell Oracle databases, middleware and applications. Catz said:

This quarter we delivered 55% gross margins on our hardware business. This is the result of the fact that really under the covers of the hardware number the Sun products are growing and the non-Sun products that are resold -- that we resell are shrinking dramatically. In addition, we are selling a lot more of the Exadata/Exalogic line. And remember that these systems are sold at good margins and also pull a lot of software with them, like [racks], partitioning and storage management.

Improving Sun operations. If Oracle just cleans up the mess that was Sun's supply chain it warrants at least a B. Sun just did a lot of things that just didn't make business sense. Catz said that Oracle thinks it can get margins near to levels when the company focused on just software. She said:

We are going to be able to ultimately bring the margins very close, if not even more than they ever were when we were just a software business. So, we still think there is actually quite a bit of room, to be honest with you. We're not quite at scale yet. We do still resell some third-party hardware. And, so we actually think there is -- it's going to continue to go up. It will obviously be higher than last year's Q4.

It's a bit hard not to be cynical about that statement. Hardware doesn't pull the margins of software---never has and never will. You could argue that if you sell less hardware that margins will naturally gravitate higher.

Nevertheless, Oracle seems to be doing better blocking and tackling than Sun did. Sun never really talked about blocking and tackling. It was either quips from Scott McNealy or a whiteboard seminar from Jonathan Schwartz. Catz added:

The first thing we needed to address was our supply chain, frankly, and getting it under control, which is now really is. And really optimizing the size that we are and not being very broadly and widely distributed all over. So, part of it is the cost, just the cost of the supply chain itself.

Getting Sun off life support. Oracle gets another high grade on this front, say an A. IT buyers know Oracle will be around. Sun was toast. There's no delicate way to put it, but Sun's future was becoming more dismal with every quarter. Co-president Mark Hurd said:

With Oracle you're certainly taking one issue that did concern some customers, which was the viability of the entity. You know, was the Company sustainable, was the Company going to be in a position? Because as you know, many of these customers made commitments and they look to be on this platform for years. And so when they make those decisions, they want to know that the Company is going to be there and be behind it. So, point one is sustainability -- I think very positive. I don't want to make it sound like we don't have work to do. This was a company that had been through several years of turmoil.

Sun's technology roadmap. On the roadmap front, the Oracle story began fuzzy and improved slightly. We'll give Oracle an incomplete here and that's probably being charitable. Oracle CEO Larry Ellison has repeatedly said that he will invest in Sparc and Solaris. On Solaris, Oracle is fighting for a flat at best Unix server pie. Exhibit A is Oracle's big squabble over Itanium with HP and Intel. Oracle last fall began outlining improvements and next releases for Sparc and Solaris.

Growth. Oracle hasn't grown Sun's business. Market share---and revenue share---is still falling. Then again, Sun is a fixer upper. Catz said Oracle can grow the Sun hardware platform. Theoretically, revenue will follow. Grade: F to incomplete. Here's a look at Oracle's hardware system revenue over the last five quarters.

  • Q4, fiscal 2010: $1.83 billion ($1.23 billion product, $598 support)
  • Q1, fiscal 2011: $1.7 billion ($1.07 billion product, $619 million support)
  • Q2, fiscal 2011: $1.75 billion ($1.1 billion product, $641 million support)
  • Q3, fiscal 2011: $1.66 billion ($1.03 billion product, $629 million support)

In Sun's last quarter as a public company---the three months ending Sept. 27, 2009---it reported product revenue of $1.18 billion and services of $1.05 billion. Those figures are a bit apples and oranges compared to Oracle's, but worth noting.

Market share. Oracle gets a C for charitable grading here. Why? For starters, Oracle isn't playing the market share game and focusing on high-end systems. In addition, Sun's market share could have tanked more. According to IDC, Sun's market share was 6.8 percent of the server market in 2010, down from 8.8 percent in 2009. HP, IBM and Dell took share from Sun. The 2011 server market share stats will give us a better feel for Sun's standing and ultimate landing spot.

idc030111b.png

Overall grade: Given the moving parts, so-so financial clarity and unknowns about future growth, I'd have to rate Oracle's purchase of Sun as an incomplete. The next year will reveal more details. Sun is better off than a year ago for sure, but we'll see if Oracle can become a hardware juggernaut. Exalogic and Exadata are a good start, but we need to see real sales figures not comments about great pipelines.

End note: You're probably wondering about how Oracle has done with Sun's software portfolio. Regarding Java, Sun never monetized it and Oracle probably won't either---unless it wins a lawsuit against Google over Android. And then there's MySQL, which was a big question mark after Oracle bought Sun. However, Oracle has updated MySQL, but the real value is control of what could be a database margin killer. Overall, it's impossible to quantify Sun's software contribution to Oracle.

Editorial standards