Last week Gartner's numbers came out and they indicated that Oracle server hardware volume dropped a whopping 40 percent in the 4th quarter and 32 percent for the year-in a market that was increasing at 17 percent. Current hardware market share has dropped below 7 percent and, in most markets, players below 10 percent are considered trivial.
In short, not only is Oracle losing market share in a growing market, it looks like that loss is accelerating, which points to why the company took a hard shot at HP and Intel's Itanium processor last month. While Oracle has clearly buried the financial impact of this failure in otherwise strong results, customers are increasingly reporting excessive pricing, which suggests that Oracle may be mining these customers for cash to overcome Sun's losses. That rarely works for long.
The problem likely comes down to two things: Oracle has lost critical mass in its hardware business and it is no longer viable, and/or Sun customers simply are not willing to ride this out and are running from the platform.
Let's explore both and close with what this means.
I first saw something like this happen years ago while at IBM when Louis Gerstner was attempting to turn the company around. He had to dramatically cut a number of units. The AS400 unit, which had historically been one of the most successful, was also one of those that had to be cut deeply to bring costs in line with revenues. The AS400 was a mid-range computer system.
When the smoke cleared, we no longer had enough skilled people to manage critical portions of the business and the lines started failing catastrophically. In that case, fortunately, people could be brought quickly back on board and the AS400 was able to continue. What made the recovery possible was that the cuts were planned, the employees were given good packages and the problem was discovered within days.
In Sun's case, the employee reduction has been going on for years with a lot of employees leaving voluntarily and being recruited for other jobs. The layoffs and staff reductions have made it almost impossible to reverse any staffing components because the critical employees either have moved on or don't want to come back.
Sun simply may no longer be a viable entity and the tools Oracle has to recover it are inadequate to the task. Making this far more difficult is Oracle's core staff, which have a software background. There are few-and I speak from experience as one of a handful of folks trained in both areas-that understand the differences well enough to translate between groups. This suggests that there are now a number of key Oracle employees with software pedigrees in hardware jobs who simply aren't qualified to do them, making even the timely discovery, let alone correction, of critical problems nearly impossible.
It was always a long shot that Sun customers would embrace a change where Oracle was their hardware provider. The discomfort many Oracle customers have with Oracle's pricing is near legendary and increasing the lock-in to a company that you are already concerned has too much control would appear to be counter-strategic to most. Instead of making these customers more comfortable with the transition, Oracle has instituted an anti-VMware position (which, I'm told, has high-price penalties for virtualized systems), which only increases these concerns, and customers have reported that they either didn't understand or believe the Sun roadmaps post acquisition, making them even more nervous.
Add to this Oracle's battle with HP, which most recently appeared to be an attempt to force HP customers using Oracle products to abandon HP, which may actually have the reverse impact. After all, HP has a roadmap that they trust and typically enjoys a higher trust score. IT managers rarely respond well in the long term to force and this latest move could backfire badly.
In short, if customers are avoiding Sun hardware because they don't trust Oracle or don't understand or trust Sun's roadmap, the fix should be on those vectors. By attacking HP in what appears to be a transparent attempt to fix falling sales in a critical area, all it will do is further reduce trust in the company and likely exacerbate its problems.
Wrapping up: Running with Scissors
Neither of these problems is exclusive and it wouldn't be the first time that an acquiring company didn't understand the business of the company acquired. Some of the biggest failures-AT&T/NCR and IBM/ROLM-resulted from these kinds of problems. However, in this case, Sun was also on life support for an extended time prior to merger suggesting that when it was acquired it may have already been dead but just didn't have the staff to announce the funeral. In short, beyond any reasonable chance of resurrection, Oracle's recent behavior may simply be an attempt to distract us from this terminal problem while it scrambles to figure out a Plan B. Until it does, this is one of those events best viewed from a distance.
Is Sun DOA? Maybe