Why IBM, Microsoft and SAP Like Oracle's Apple Strategy With Sun

Earlier this week I got word that Oracle was ceasing the Sun Solaris licensing deals with other hardware vendors. This reminded me a great deal of what Steve Jobs did when he returned to Apple and killed the Apple clone business.

In both cases it reflects a realization that a company cannot effectively create competition for a prime business. In Apple’s case, Apple was a hardware company and had no business even trying to be a software company unless it was willing to give up hardware. Sun’s last CEO, Jonathan Schwartz, seemed to be trying to turn Sun into a software company but his effort stalled someplace in the middle and Sun foundered.

Oracle is clearly a software company, yet it seems that with the Sun acquisition, suddenly they seem to think they are Apple -- but they aren’t. While Oracle is certainly much better run than Sun was, and Apple was before Steve Jobs return, wouldn’t they have as much trouble transitioning to a hardware model as Sun did going the other way?

On the other hand the desktop side of the technology industry is certainly moving to tight vertical integration and Cisco seems to be on a similar path on the IT side of this market. Maybe Oracle is onto something. Let’s explore this.

Sun’s Problem

Sun had a leadership problem, Larry Ellison is pretty hard on Jonathan Schwartz but he really failed to point to the cause of the issue, which was that Jonathan was the wrong guy to lead Sun. Schwartz was a software guy and Sun was a hardware company.

If you’ve ever worked with both groups you’ll quickly learn that they don’t really talk the same language. They rarely get along well. You take someone with a particular expertise and put them in to lead a company and they will likely try to turn that company into a form they recognize.

Schwartz tried to turn Sun into a software company and started licensing out the key technology, much like Microsoft does very successfully. But he needed to transition the company from one model to the other quickly. And if your followers can’t understand you and you don’t understand the major portion of your existing business then you won’t be successful. And saying Schwartz failed is a gross understatement.

The only other way to do this would be to spin out the software unit as a different company. This worked better for Palm initially but the end result for them was that both units eventually failed. This was partially because they waited too long and the market had moved away from PDAs and partially because first software and then hardware executed poorly.

Apple went the other way and stopped the transition into a software company and doubled down on hardware. However, they likely might have still failed if it hadn’t been for the iPod which was enough of a success to return the firm to success and profitability.

Still, the iPod was a hardware product and this would indicate that the most successful path would be to retrench as a hardware company, and look for other hardware products that the market wants and not switch focus to software.

Vertical Integration

I’m calling vertical integration one of the big trends for this decade. We’ve seen Microsoft increasingly drift over into hardware, most recently with the Kin phone, HP just picked up an OS with the purchase of Palm, and Apple has started to actually design their own chips with the A4 used in the iPad.

Cisco just bought their own design house (Moto) and has been building an IT focused vertical around the Cloud (servers + networking) for awhile. This is starting to look like a game of musical chairs where the company that builds a horizontal stack that is left without a dedicated vertical once the music stops will be the loser.

In a way this is feeling like a return to the IBM years, where you would go to one company for an entire solution.

But Oracle’s goal appears to be to create Apple-like products for the data center in terms of data appliances that are effectively plug and play. Granted they will likely be assembled to the specifications of the IT organization that buys them. But Oracle likely will maintain the result and because they are relatively standard, this could result in lower acquisition and maintenance costs for the buyer and higher margins for Oracle.

But this doesn’t come without risk.