CFD3 Prep Post: Oracle

This is part of my series of posts on Cloud Field Day 3.

Firstly let us marvel at the fact that Oracle is appearing at Tech Field Day in the first place. This is an amazing thing.

Oracle has made an art-form out of never talking to anyone who isn’t a current customer and Larry Ellison is famous for saying “The interesting thing about cloud computing is that we’ve redefined cloud computing to include everything that we already do.” Among other things.

Oracle wasn’t first to $10 billion in revenue, Salesforce was, and no one (outside of Oracle) believes Ellison’s pontificating about how it’s totally a cloud leader.

To see Oracle appearing at Tech Field Day is pretty amazing in this context.

Oracle Buying Spree

Ellison is not afraid of spending a lot of money to keep his company relevant. Yes, technically Oracle is run by co-CEOs Mark Hurd and Safra Catz, but Ellison is still executive chairman and he’s the largest stockholder. It’s his company.

Oracle bought NetSuite in 2016 for about $9.3 billion. It’s a SaaS ERP system, used by big enterprises.

Oracle also bought Dyn (network management and DNS) for $600 million and Ravello Systems for $500 million. Last year it bought Wercker which is a continuous integration platform, largely, though we don’t know how much it paid. It also bought Aconex, a SaaS tool for project management largely used by the construction industry, for $1.19 billion.

It’s largely a mix of cloud-ish software things and SaaS services that big enterprises use. Why? Two reasons.

Why The Buying?

It’s faster to buy than to build, and Oracle doesn’t have time to build these things to catch up. It ignored the cloud market for waaaaay too long and now it has to spend big to stay relevant to a world where cloud is totally viable for big and important applications and big organisations with lots of money are happy to start giving it to people other than Oracle.

Oracle has long known that it’s the enterprise applications and the data inside them that matter, and the Oracle database is at the heart of everything Oracle does. Applications built on top of Oracle’s database are incredibly hard for enterprises to move away from, which is why they keep buying it even though lots of customers resent giving Oracle money.

A note of caution though: there are lots of DBAs and customers who love Oracle, because they are enterprises who just want to get stuff done and Oracle has been, for a long time, a really good database. Large organisations have to deal with a lot of stuff that is boring and complicated, stuff that relatively new startups often haven’t worked out how to do yet, and may never bother. Just call it legacy and move it. Just replace it, you big squares!

New apps just aren’t built on Oracle, largely because the entry price is too high. If you’re a startup, MySQL or PostgreSQL or SQLServer or MongoDB or any one of dozens of other options are plenty good enough. And some of those startups (Facebook, Instagram, Netflix) got really big and are doing big and impressive things with non-Oracle databases. That has shown big, serious enterprises that they can run big, serious applications on top of these non-Oracle databases.

That’s an existential threat to Oracle, and it’s one that AWS is more than happy to exploit.

Oracle is racing to convince customers that it isn’t legacy, it’s proven and that they should care enough to pay the premium for Oracle software and cloud.

How’s The Pace?

At the Ravello Blogger Day last year, Oracle started to show some of its new open side by having Clay Magouyrk, VP of Oracle Cloud Infrastructure, present to us in a refreshingly honest and open way about the state of Oracle’s cloud infrastructure. As my friend Chris Wahl wrote at the time, Oracle’s vision was to provide something usable for the many enterprises with lots of applications that they don’t want to have to rewrite just to get some cloudy goodness.

I’ve told the same story myself: 80% of writing software is putting in bugs, and the other 80% is taking them out again.

Enterprises don’t have the budget to rewrite a bunch of apps that are “working fine” as far as they’re concerned, and they have experience with how often IT projects go wrong. It’s easier to just leave them alone, or re-platform them onto something else. This was a major driver behind a lot of the early VMware adoption and Windows NT apps getting moved onto new hardware to buy an organisation time. You can get a lot of improvement by just throwing hardware at the problem.

The theme continued at this year’s Ravello Blogger Day (that I sadly couldn’t attend) with Oracle using the phrase move-and-improve instead of lift-and-shift. Oracle is making a dual play here: keep existing customers on Oracle stuff even if it gets moved to ‘cloud’, and try to appeal to developers writing enterprise applications that Oracle is a reasonable stack to write apps for.

Is It Working?

I have opinions on whether or not this plan is advisable, and while Oracle isn’t in the same league as the hyperscale cloud providers, it is a force to be reckoned with in enterprise data management and ERP type apps.

Most of Oracle’s revenue is still software licenses and product support, and overall growth is flat. Its cloud competitors are also spending big, and they also have plenty of money to slosh around and a willingness to do so. Ellison won’t outspend them.

I look forward to seeing what Oracle has to say about why people should be excited about its products and technology. Hopefully the new trend of openness and willingness to seek out new customers will help us learn a bit about the plan.