Why Oracle May Really Be Doomed This Time
I was laying awake early this morning thinking about Oracle.
Back when it was starting its $20 billion buying spree of companies in 2006, analysts all said the strategy would fail. There were a lot of compelling arguments why.
People said that the assets of a software company are the engineers and the engineers would go elsewhere. People argued that Oracle fundamentally didn't know how to write good application software, and they couldn't buy their way into changing that DNA. People argued that stitching together the disparate systems of PeopleSoft, Siebel, Oracle into Project Fusion would create a Frankensteinian mess of unshippable bloatware. I once wrote a BusinessWeek profile of John Wookey-- the man tasked with Fusion in the wake of all the purchases-- called "The Hardest Job in Silicon Valley." It's telling that he's now at Salesforce.
At the time, I was far more bullish on Oracle's chances than all of these naysayers for a few reasons. First off, a horrible business application was all relative. The entire category sucked. Oracle's applications were worse than SAP's, but no one relished using SAP either.
PeopleSoft and Siebel were far better with their so-called "best of breed" applications that focused on HR software and customer relationship management. But even they were too hard to use, configure and required employee training. There were famously millions of seats of Siebel software sold in the late 1990s and early 2000s that were never used.
Second, Oracle had nailed the hard stuff: The underlying technology layer. SAP simply couldn't compete on database or middleware technology. There was so much pent up anger that companies had been forced to spend billions on consultants to make enterprise software work together. By the mid-2000s there was a huge push to streamine things to one vendor: One person to call and scream at if things went wrong. One vendor that couldn't pass the buck. If you were going to go with one vendor, Oracle was simply the better option because the underlying technology was just better and that's the stuff that had to work.
The third reason was that Larry Ellison just got what the market wanted at that moment, and he played it perfectly. He knew that this was a moment in time in the software application wars where the best product simply didn't matter, because everyone had resigned himself or herself to the fact that enterprise software just chronically under delivers on what the sales guy promises.
What customers wanted instead was convenience and stability. So much work and money had gone into implementing these programs, and no one wanted to rip them out, no matter how much they hated them. They would happily just keep paying on-going maintenance fees to stick with the status quo. By mopping up the cash-rich but growth-challenged enterprise giants, Oracle could amass an installed base and bottom line that would be the envy of the tech world.
Even a lot of the employees stuck around longer than analysts expected. Because there just weren't much better alternatives in an enterprise world that was becoming a two-horse race between Oracle and SAP on applications, and a two-horse race between Oracle and IBM on the database front.
The wild-card was the looming threat of software as a service and open source software. But while both were superior options for Web business and smaller companies, both were seen to be risky bets for large, staid organizations. Ellison saw the promise of SAAS-- that's why he personally invested in Salesforce and Netsuite. But he knew Oracle didn't need to make any knee-jerk reactions because companies didn't want innovation back then. They wanted predictability and stability.
The best product didn't matter.
The reason I'm suddenly turning less bullish on Oracle is that I believe that is finally changing. There are a few reasons why:
1. Time: Companies can put off installing new technology for a long time, but not forever. We are finally reaching the point where enough time has passed that the wounds of those horrible 1990s software implementations have healed. CIOs know they can't keep those stodgy old systems forever.
2. Millenial entitlement: The reason Oracle's rollup strategy won was because companies were pummeled into a place of Stockholm-syndrome-like acceptance. When I was covering Oracle in the mid-2000s, I spoke with customer-after-customer, and I can't say any ever loved the software they spent millions on. No matter what vendor they used. Most just sort of threw up their hands and said that so many salespeople's promises had been broken over the years, that they were just resigned to the reality that no enterprise applications would be easy, intuitive or do what they promised. It's kind of like cable companies. We just know they suck, and we don't really expect more.
But this is one place where Millenial entitlement is a good thing. Millenials are coming into the workforce and the generation has an amazing capacity to demand the world revolve around their desires, whether that's reasonable or not. Millenials will just start demanding better software from the companies they work for, and if they don't get it, they'll start installing their own skunkworks implementations.
3. Technology: The technology underlying cloud-based systems has gotten more sophisticated, and computing itself has changed with the rise of iPhones, Androids and iPads. Applications can do things now that they just couldn't in the late 1990s.
As a result, some of those early promises of application software that failed, may actually be possible. Take collaboration software for instance. Billions have been spent on horrible collaboration software, but for many teams something as simple and elegant as Yammer running on a browser or a phone just works. Ditto, Jive and Asana. It actually does make teams collaborate better and lessens the crushing load of email, in my experience.
Now, these tools don't work for all workplaces and getting into real enterprise level will still be a long land-war. Longer than most people expect right now, I think. But these younger companies are starting with what technology can do now, versus what it could do a decade ago.
4. Options: This ties into time and technology, but the options on the market are finally starting to get substantially better. So much better, that employees who work at a company with newer tools will balk if they go to a new job with olders systems. CIOs are finally starting-- starting-- to see that truly intuitive "consumerized" technology that employees will actually use may be worth the pain of switching.
Any single one of the reasons wouldn't have been enough to turn the tide, but taken together, it's a watershed moment for enterprise software. Not only could it dramatically change how we all work, but it could translate to billions and billions in returns for Silicon Valley just as we enter a lull or mini-shakeout in consumer Web sites. Because enterprise adopt technology so much slower than consumers, I'm increasingly convinced we are at the cusp of something that only comes around once every a decade or two.
Typically I think big public companies are inept, but I would not count Oracle out. Ellison has long shown he viscerally gets where the stock market, the customer and the technology are going. He may be better at this than anyone leading a technology company today. He has pulled off stunning and dramatic turn arounds of Oracle in the past. He can force the company to shift out of sheer force of will, uncowed by the near term pain he may inflict on customers, employees or Wall Street in the process. That's something that only Steve Jobs and Jeff Bezos can do well.
But to win, Oracle will have to change its strategy as dramatically as it did in 2006 when Ellison famously announced that software innovation was dead and just started to buy everything. Buying once-hot companies like Taleo and RightNow isn't going to cut it this time when there are better products in the market like Workday and Salesforce.