Mark Hurd, the CEO of Oracle, is a numbers guy. In a recent CNBC interview, he was asked about competition from high-flying MongoDB. Rather than respond directly to the challenge, Hurd said, “Just look at the numbers and look at the facts, and see what they're telling you.”
The potential good news for Hurd is that those numbers and facts may keep Oracle comfy during its dotage. The potential bad news is that those same numbers and facts suggest Oracle has lost its way as a general-purpose database. Oracle is, in short, seems to have become the database CIOs will pick to run their business, but not to define their business. Is that a bad thing?
Oracle lost the developer war but won the marketplace battle
Put another way, how is it that Oracle can be such a miss with developers and yet still print billions of dollars in revenue? I’m guessing Hurd doesn’t care much about developer adoption. He likely doesn’t care that, measured in popularity, Oracle has been in terminal decline for many years, as DB-Engines’ data suggests. In fact, if you look at the database technologies for which developers ask the most questions (indicating production use) on Stack Overflow, only MongoDB and PostgreSQL are booming (of the top-five database technologies).
No, what Hurd cares about, as he acknowledges in his remarks to CNBC, is the fact that Oracle still controls roughly half of the global database market, worth tens of billions of dollars. Never mind that, as Gartner analyst Merv Adrian has highlighted, Oracle has lost market share every year since 2013, and collectively the old guard relational database players have shed nearly five percentage points.
Given how entrenched they’ve been for so long, that still leaves them with roughly 86 percent of the market, but it’s not good news for them that open source databases, which claimed 0 percent of the paid market ten years ago, now take more than 7 percent, according to Gartner, or that cloud databases from Amazon Web Services and other new commercial providers are also booming.
Developers are driving those open source and cloud database trends, but they’re still not up-ending Oracle’s database reign. Hurd can perhaps afford to be dismissive of MongoDB’s $250 million in revenue last year, and even of its 400 percent growth rate in its database-as-a-service business. Why? If cloud and open source are such a big deal, and if developers can’t be bothered with Oracle (and even seem to prefer to get its MySQL open source database from AWS), why is Hurd so confident?
Because … applications?
Oracle is all about the job of running the business
I don’t know that Oracle believes its database reign will ever end, but years ago the company didgrasp the need to extend its dominance into applications. As open source developer Paul Ramsey has noted, “I'm not even 50 percent sold on the idea that Oracle's performance in the core [database] space matters much to them. The gold in them thar hills is the next level up: Oracle Financials, HR, etc. The lock-in that will squeeze customers hard for the next 25 years.”
Developers, not CIOs, increasingly bring new technology into the enterprise. Well, not all of it. They may bring in new technology for infrastructure used to build applications, but they don’t drive the technology decisions for the boring, run-your-business applications that Ramsey spotlights. Developers are not test-driving Oracle Financials. Instead they’re building the applications that will give an enterprise a competitive advantage, not those that calculate employee growth.
To be sure, these “run-your-business” applications are critical to operating an enterprise at scale, and with so few options for CIOs to choose Oracle will be kept in a smoking jacket and slippers throughout its old age.
In this way, Oracle can afford to lose the developer war so long as it’s prepared to take on the comparatively small but very profitable role of being the company (like SAP) upon which enterprises run their business. It won’t make Oracle popular, but it will make Oracle billions.
In other words, Hurd is likely correct that the facts and numbers support his contention that Oracle is in a safe place, with “safe” meaning “we’re extending our database dominance into a period of dominance in run-your-business applications.”
Those same facts and numbers indicate that what Oracle is giving up is its old role of defining the data behind customer engagement. That new world is moving to new players like MongoDB and AWS, as well as established competitors like Microsoft. You may well find that new world more interesting (I certainly do), but for Hurd and Oracle, “interesting” and “popular” aren’t the drivers behind its decision. “Billions” just might be.