I've been getting calls all day about President Obama's visit to Silicon Valley. One of the folks he is visiting is Larry Ellison. I found this interesting because Larry is probably the closest thing we have to a politician running an enterprise-class company. Even EMC has been making fun of him of late with some cute take offs of the old Mac vs. PC ad campaign that Apple did. What both IBM and EMC are clearly having issues with is Oracle's over-the-top claims about his technology. While EMC has fun with a video suggesting that Oracle is "artificially enhanced," IBM seems to come right out and say he cheated on his system benchmarks.
Let's talk about the practices of politicians and business leaders this week.
For some time, I've believed that President Obama could learn a lot from Steve Jobs. As part of this trip, he is meeting with Steve. Steve Jobs does one thing particularly well, and that is presenting a product in a way that people become convinced it is magical. This skill is critical to a successful politician who needs to capture the imaginations of constituents in order to assure not only the passage of an unpopular budget but a re-election. What is particularly interesting is that Obama is meeting with executives from Twitter and Facebook, which could be instrumental in his reelection bid, showcasing, perhaps, the perceived power of these social networks.
But if the message isn't improved, making it reach more people will only turn them against him. That was what happened with what has been renamed Obamacare. He lost control of the message and now this massive accomplishment stands more as a liability to his continued presidency than it does as an asset. It's always quality over quantity, and mixing those two up is a key reason why the first two years of the Obama presidency are perceived as being unsuccessful.
Ellison vs. Jobs
One of the differences between Ellison and Jobs is that while both set incredible expectations that are almost impossible to achieve, Jobs more consistently drives his folks to achieve them. He can showcase a product, like the first iPhone, that can't possibly work and drive his folks to get it working in a few short months. And he does this regularly. Ellison has a tendency to make big promises (remember his thin client claims), but miss on execution. Oracle carries nowhere near the customer satisfaction that Apple enjoys.
In addition, Steve knows better than to over-promise on a future product. All that does is reduce his credibility and get people to defer purchases. Oracle evidently doesn't think the same way, because it recently announced that along with Fujitsu, it would be increasing performance of its Sun-based systems 15x over the next three years.
As luck would have it, when I read this, I had just finished a meeting in which the analysts had given NVIDIA grief for suggesting it could improve performance 5X in under two years. For processors, we typically think 20 percent performance increase year over year is the high bar in terms of advancement and forgot that graphics companies like NVIDIA have a vastly higher bar, often increasing performance 2x or 3x. But even for them, 5x is a stretch. The only thing making this possible is that it is a System on a Chip from a company that regularly achieves what it promises. Oracle isn't a graphics company, and this isn't a System on a Chip, leaving me with the feeling that either the existing systems are incredibly slow or it has suddenly stepped into Oracle's version of the reality distortion field and can't get out.
This feels like Oracle trying to obfuscate its lack of a strong roadmap with crazy promises. The big problem for it, though, given the life cycle of large systems, is that a 15x performance increase will clearly make obsolete any other system you have from the company. The advice this would drive from analysts would be to hold off buying anything more until the company can demonstrate this capability. Then you'd order that. Otherwise, you'll be stuck with a bunch of obsolete hardware and have to not only write it off but explain why you didn't get the implication of Oracle's announcement.
A Joke to Remember
My grandfather used to tell this joke. A guy goes into a tire store and asks what the dealer charges for tires for his car. When the dealer says $25, he responds, "But the guy up the street only charges $20." The dealer responds, "Well, then go buy them from him." The guy says, "I can't, he doesn't have any," to which the dealer says, "Oh, well, when I don't have any tires to sell, I charge $15." The point is, it doesn't matter what claim someone makes if they don't have anything they can sell.
In the end, if you believe Oracle, and I wouldn't, you'd be best off waiting to buy from the company when these new superfast system arrived. If you don't believe Oracle, then it isn't trustworthy, and you'd be better off not buying from a vendor you don't trust.
And that is why I think making claims like this is a stupid practice.
Wrapping Up: Vendors and Politicians
In the end, the messages that vendors have about something they will do in the future, like those from politicians, should be taken with a lot of healthy skepticism. Both parties have a tendency to over-promise and under-deliver. However, the most successful actually execute on amazing promises and that goes to the core of why they are so successful. Oracle made a three-year promise for a product that has a useful life that could extend to a decade.
Let's hope Obama listens to Jobs more than he does to Ellison, otherwise he'll likely not be around as president to see whether Oracle can get a 15x performance increase.