Why Oracle won't beat Amazon

Sure, Oracle has made serious inroads in enterprise SaaS. But the company's chances in IaaS remain slim

Why Oracle won't beat Amazon

Let’s give Oracle’s Larry Ellison some credit for bravado. As captured by Eric Knorr, “Oracle is making it abundantly clear that it's not ready to concede anything to the cloud natives” like AWS.

That’s awesome! Or it would be, if only Oracle could make this “abundantly clear” to customers, who remain mostly oblivious of Oracle’s heroic attempts to dominate the public cloud in the same way it has dominated datacenters.

In the SaaS world, Oracle is a player. (Note: Matt Asay is vice president of Mobile for Adobe, but his opinions are his own.) But in the explosive IaaS market, “Oracle also does not have enough market share to qualify for inclusion” in Gartner’s Magic Quadrant. It’s a non-entity, and despite a year of bluster that hasn’t changed. 

Unfortunately for Oracle, this will only get worse. Oracle’s bread-and-butter large enterprise customers have told McKinsey they expect to jump from 10 percent public cloud adoption to 51 percent by 2018. Unless Oracle does something drastic -- soon -- that momentum will continue to feed the cloud natives that Ellison disparages.

I think I can, I think I can…

This week is Oracle OpenWorld, an annual rite in which Oracle declares that every competitor is toast.

Part of Oracle’s would-be cloud advantage, according to Oracle CEO Mark Hurd in an interview with InfoWorld, is its ownership of the full cloud stack. That is, Oracle can fine-tune everything from software to silicon to deliver “a very large advantage in terms of the depth of our knowledge and our technology at each layer of the stack.”

The problem with this line of reasoning: It’s simply untrue.

A few years back AWS kernel and operating systems chief Chris Schlaeger told a LinuxCon crowd that “the ability to modify every parameter of your system is key,” and has been standard operating procedure at AWS for years. The company doesn’t buy stock hardware from anyone, and instead heavily modifies all aspects of its datacenters -- from software to hardware and, yes, that includes silicon. In other words, Oracle isn’t leading the pack with its full stack experience. It is trailing.

Even so, Knorr is correct to argue that “Oracle's cloud push cannot be dismissed.” Given the stakes, Oracle cannot afford to lose the cloud race.

Accelerating the cloud

That race, incidentally, is heating up.

According to a new McKinsey survey of more than 800 CIOs, investment in public cloud resources is booming. Among the large enterprises that can afford Oracle, the trend is particularly pronounced:

cloud computing

These same companies are investing heavily in private clouds, but the biggest shift is in how they view public IaaS, where AWS reigns (by a huge margin). Private cloud is a transitional phase, a stopover for CIOs as they grow comfortable with public cloud. As the survey data reveals, however, that comfort level is growing -- remarkably quickly.

Meanwhile, investments in traditional datacenters are in sharp decline, which shows in both the McKinsey data and in Oracle’s earnings. As Host Analytics CEO Dave Kellogg calls out, “Oracle continues to struggle, missing revenue consensus 12 of the last 15 quarters and non-GAAP EPS 6 times of the last 15.” In the latest quarter, the company saw new software licensing revenue plummet 11 percent.

On the positive side, Mark Hurd told InfoWorld that the company’s cloud, SaaS, and PaaS revenues have grown 82 percent year over year. While most of that growth is SaaS and PaaS, not head-to-head with AWS in the IaaS market, it’s still important. The question, however, is whether the company can manage the transition to cloud even as its traditional business flattens and falls.

Oh, and whether it can be one of the few clouds CIOs are going to invest in. By Hurd’s own admission, “The customer wants one-stop shopping as much as they can. I don’t think customers are going to have one cloud, but I don’t think most big customers are going to have 10.” Maybe, he tells InfoWorld, they’ll have three. That means that Oracle has a heck of a lot of work to do merely to get past Google and Microsoft, never mind AWS.

Impossible? No. But Ellison is going to need more than bold keynote declarations to win. He’s also going to need more than fancy engineering, which his top-three IaaS competitors have in spades, and arguably can do significantly better than Oracle. Instead, Oracle will need consistent execution over many years -- plus a heavy dose of luck -- to recover the ground it lost by sitting out the early cloud market as “vapor” ware.

Copyright © 2016 IDG Communications, Inc.