The Blog

What the #&^$ Is Happening at Oracle Cloud?

Calendar Icon 06.24.2020
aws-section-divider aws-section-divider

There’s something going on over at Oracle Cloud, and I think the industry is missing it entirely in favor of the easy “Oracle sucks” narrative.

There’s a change afoot, and I have a sneaking suspicion of what it might be.

True, Oracle Cloud is a popular target for derision and mockery—and who could blame anyone? Oracle is a multi-billion dollar company with a long history of suing its customers, getting sued by its customers, threatening to sue its customers, having a chairman whose statements are seemingly untethered to reality in any meaningful way, and other various points of contention.

Also true, last week’s earnings report showed that their cloud division (which inexplicably includes license support) grew only 1% year-over-year.

And true, I did dip into the Duckbill Group’s Spite Budget and make a beer commercial mocking their chairman’s 75th birthday last year.

Every story needs a villain, and Oracle (and by extension, their chairman) has served that role admirably for the four years I’ve spent in the cloud industry.

This has all happened before

In 2014 (and much more noticeably, in 2016–17), Microsoft began doing some very strange things. They made a bunch of fascinating hires and acquisitions. (For a time, it felt like every DevReloper in the advocacy community was suddenly working at Microsoft.) They announced that some guy named Satya Nadella would become their new CEO. Soon after, they renamed “Windows Azure” to “Microsoft Azure.” And the entire industry basically laughed them out of the room.

The industry reaction is very different today. Azure is a serious competitive threat to AWS in the context of “corporate IT moving to cloud.” They’ve got a lot of incredibly talented people, a 40-year history of speaking directly to the enterprise in terms that everyone understands, and they clearly grasp that the cloud is the future of Microsoft—whether they want it to be or not.

What I see with regard to Oracle Cloud feels an awful lot like the beginning of that evolution.

Follow the people

If we look at some of the folks who have left AWS to go work on Oracle Cloud over the past year or so, you start to see a few interesting points emerge.

These are high-level people, by which I mean “multiple VPs who have been largely quiet about their departures for reasons I can probably guess and will respect by not naming them here.” These are executives who tend to Think Big™ in ways that Oracle historically hasn’t exactly championed.

It’s rather clear that they’re dreaming of something bigger than “build a ‘cloud’ to host a bunch of boring enterprise workloads.” You might be able to make an exception case for Ariel Kelman (now Oracle’s Chief Marketing Officer) wanting to move to a company that de-emphasized “frugality” in favor of other approaches, such as “putting our name on a stadium,” but you’re not going to be able to dismiss some of its recent hires so easily.

Something compelling must be afoot to attract some of the folks they’ve been able to hire—folks who focus on very modern technologies that don’t align with the mental boxes people generally slot Oracle into.

I might be wrong

This hypothesis isn’t without its problems. Most obvious and painful is Oracle’s lack of serious investment in data center buildups.

Last year, Oracle invested less than $7 billion in CapEx (Edit: WAY less than that! Just over $1.5 billion; their filings show quarterly breakdowns of “trailing four quarters” to trick people into making exactly this mistake. Thanks to Charles Fitzgerald of Platformonomics for pointing this out!) into the business while returning nearly significantly more than that to shareholders in the form of stock buybacks and dividends. That’s not going to lead to cloud greatness in an era where the top three spent almost $73.5 billion in 2019 alone.

The best software in the world built by the best engineers still needs hardware to run on. And that hardware needs to live in data centers close to where their customers are. If you want to have a thoroughly built out region in 2020, you needed to begin building it several years ago.

Oracle also is in use by basically every public company out there, less a few notable exceptions. The challenge remains: Is this sufficient? Is there a growth path for continuing to expand into new markets and grow these customers or is this a 40-year “long-tail” of a dying industry position?

The latter “feels” right to me, but that’s shooting from the hip. It may well be that either I’m wrong, or that this is a vision on a long enough time horizon that Oracle doesn’t really care.

Oracle’s greatest challenge

It’s clear that at least someone inside of Oracle Cloud has their eye to the future, and aligned with the rest of the world’s vision of public cloud.

Oracle Cloud’s single greatest challenge is and is likely to remain the word “Oracle.” Their name is effectively mud in the developer space, and it’s precisely that constituency who determines which cloud provider many companies choose for their workloads.

Their second (and I suspect closely related) greatest challenge is, of course, the lack of internal political will to set Larry adrift on an ice floe so he stops making ridiculous statements that nobody actually believes.

While it’s easy to say that these add up to an insurmountable challenge, look at Steve Balmer and Microsoft. There was a time where every point I’m making would have applied to them, and you would be laughing at me for taking Azure seriously.

Nobody’s laughing at Azure now. In fact, some are whining in increasingly insecure ways about them.

My point is not to suggest that Oracle Cloud is somehow going to be the dark horse cloud provider that everyone should be fearing; that would be premature.

But I’ve used aspects of Oracle Cloud and found them surprisingly well-built and easy to use. I see who they’re hiring and recognize a potential future that may come to pass.

Of course, it also may not.

But I’m not counting them out just yet.

aws-section-divider