Mark Hurd and the Sun/Oracle Losses

Bill Moran's picture

September 6, 2011

 

It seems reasonable at this point to examine how Sun is doing, and what its prospects might be. Many Sun customers of course will be very interested in this subject since they need to make plans for future installations of Sun hardware.

 

First, let’s review some of the factors that created the environment before Hurd joined Oracle. The Oracle/Sun deal closed at the beginning of 2010. By that time, many Sun customers had become uneasy about Oracle's plans for Sun. Not only were the customers uneasy, but many Sun employees were looking for the exit. Given the Oracle statements and plans to increase the margins on the Sun business to make it more profitable (or, perhaps, to bring it up to profitability), they (quite naturally) assumed the result would be many layoffs when Oracle took control. Other employees probably decided that they did not want to work for Oracle for various reasons. In all, we believe this led to a loss of technical talent, especially around Sun’s SPARC microprocessor, and a loss of hardware sales talent.

 

In addition, it is very likely that Oracle had underestimated the difficulty of managing Sun’s hardware business. Hardware company operations differ significantly from a software company. For example, software companies typically lack a manufacturing arm with all the planning, logistics and interdependencies that implies.  Also, the breadth and depth to a software supply chain is nowhere comparable to that of a hardware company.

 

Historically speaking, Sun had actually only entered the large server business by acquiring Cray’s server business. The author of this piece attended a Sun analyst conference several years after that acquisition At that event, Scott McNealy, Sun’s CEO, confessed that he was only just beginning to understand what the large server business was all about. Remember, Sun had been a workstation company long before the Cray acquisition, so they were familiar with many of the requirements of the hardware business. In spite of this background, it took Sun a relatively long time to develop and bring to market a new high end server to replace the machine they acquired from Cray.

 

Turning to Oracle, we see little evidence that the management had any great understanding of the differences between the software and hardware business. One would naturally expect some stumbles and some did occur – Oracle’s disappointing hardware results last quarter (down 6%) are an indication. Hurd has his own set of baggage and preconceptions including a reputation as a cost cutter very oriented to financial results. His time at HP was marked by large layoffs and pay cuts. This was unlikely to reassure Sun employees concerned about their future. While, one could agree that some of this is probably needed at Oracle/Sun, other management skills are also necessary. For instance, one skill necessary for hardware management is some engineering understanding. Hurd was a salesman and manager. He lacks a technical background, and his experience at HP would probably not have given him any. As the HP hardware division was performing reasonably well for him, he did not need to create or develop that business – rather, he was focused on cutting their costs.

 

Oracle initiatives to revive Sun have focused on benchmarks and ads promoting Sun’s performance when they outperform HP and IBM. More often than not these benchmarks involve esoteric hardware configurations. Recently, they have focused their fire on Teradata. Such a strategy is not likely to appeal to and retain existing Sun customers more concerned about pricing and support issues. As evidence, Oracle has been gaining in profitability as a software company but losing market share in the server business. 

 

What conclusions to draw? First, it will take more time for the management team at Oracle, including Hurd, to learn how to successfully manage the hardware business along with the challenges of an integrated software/hardware business. Oracle/Sun customers should expect more bumps in the road. Next, users need to decide if they’ve got confidence that Oracle, with its very heavy software focus, will do what it takes to grow their hardware business and return it to ongoing profitability, or are they just dabbling in hardware to help sell software licenses. And if it’s the latter, the question is how long will Oracle be able to sustain, and invest in, a shrinking hardware business?

 

By Bill Moran,

VP Research, Ptak, Noel & Associates

*Mark Hurd has been at Oracle for almost a year. He left HP one year ago and joined Oracle in September of 2010. He was brought in specifically to work with combined hardware and software products.