Oracle Makes Its Case to Move Legacy Mountains to Cloud

eWEEK DEEP DIVE, PART I: CEO Mark Hurd is the chief taskmaster in persuading old-line, institutional Oracle customers to move their humongous billing, ERP and other well-entrenched systems to a new paradigm that will take them safely ahead in the 21st century.


SAN FRANCISCO—If you didn’t know Mark Hurd by sight, you might think he is the premier of a sovereign nation. At this week’s Oracle OpenWorld conference, the sharply besuited executive was escorted wherever he went by some rather large bodyguards, and he travels—even if the destination is a block away—in a large and serious black SUV, followed by his handlers in a similar vehicle. If a place he is to meet someone somehow doesn’t look airtight secure, people start talking into their wrists, and the location is immediately changed.

While he doesn’t exactly run a country, Hurd, 61, is one of two CEOs of Oracle Corp., one of the oldest and most successful database and enterprise IT providers in the world for more than 40 years. The New York native and once-upon-a-time Baylor University tennis star is said to have an effective ying-and-yang relationship with his counterpart, Safra Catz, who’s also CEO. Generally, she handles the numbers, investors and Wall Street; he handles the marketing, product development and sales.

Both represent the company to customers, investors, partners, the media and heads of state. The duo serves at the pleasure of their Chief Technical Officer, Chairman of the Board and company co-founder, Larry Ellison, and together they run a company with 140,000 full-time employees globally.

Company Has Been a Survivor

Oracle, whose futuristic-looking headquarters location was once a Marine World theme park on the San Francisco Bay in Redwood Shores, Calif., originally hung up its shingle to sell, install and maintain parallel databases for enterprises in the 1970s and ’80s. The company still does that and a lot more here in the sunset years of Decade 2 of the 21st century.

But the IT world has changed a wee bit since the company began creating columns and rows in structured storage in 1977. The House That Larry Built has managed to remain relevant all these years by continuing to be a core system component and by innovating; it is a true survivor in the IBM sense of the word—although that big competitor (AWS and SAP are two other major ones) has quite a few more years on it than does Oracle, having been born in 1916.

Hurd has been president or CEO of Oracle since September 2010, after he resigned after five years as CEO of Hewlett-Packard Co. Within weeks of his departure, Ellison called up his friend and offered him the Oracle job, knowing his personality would fit well within the culture. And it has.

Hurd has supervised some of the most profound product changes and updates in the company’s history as Oracle continues to move more and more of its software into cloud subscription formats that are totally different from its traditional on-premises-based applications for servers. He came aboard immediately following the momentous $7.4 billion acquisition of Sun Microsystems, which put Oracle into the data center hardware business of servers, storage and networking—segments that he had run while at HP.

A Fair Amount of Convincing to Do

Now Hurd is tasked with convincing a lot of old-line institutional-IT customers (such as government agencies, the military, multinational corporations, aircraft makers, oil and gas developers, scientific labs and others) that they need to retool their vast IT systems, move important applications into distant data centers that are not their own, and trust companies like Oracle and its partners with their crown jewels: data.

Not an easy assignment, this. Think about it: All these huge companies have been investing millions (often billions) of dollars in buying new and upgrading old hardware and software and services, and what’s being used on a daily basis is often older but still efficient servers running older but still efficient software for employees who have long since figured out all the shortcuts and know the strengths and foibles of all this stuff intimately. And the older apps have been customized for certain people and layered with new functions over spans of years. This doesn’t translate to the cloud well at all, because all that customization tends to get lost as a system becomes transformed into new-gen production ware. That’s all code written for specific servers, networks and security purposes.

Like the home handyman who shows you what part of the dishwasher to tap with a rubber hammer when it decides to stop working, veteran employees know how to finesse old software to do what they need it to do. Assigning them to a whole new way to use IT is often fraught with push-back, and for good reason: Work doesn’t get done, customers aren’t served and people don’t get paid when processes get hiccups.

How He Sets His Approach

How does Hurd go about doing this? The pressure’s on for sales of cloud apps and services to happen, and happen quickly. This means lots of old processes will be going away.

“I think your point is correct in that these (older) applications have been customized to agree with the processes of the company, as opposed to the processes being changed to fit the software,” Hurd told eWEEK in a one-to-one meeting following his Oct. 23 keynote at OOW. “What we can do today, since most of these systems were built 20 or 25 years ago, we can go in to a customer with a tool play—such as Business Suite, PeopleSoft or J.D. Edwards—we can go in with a set of tools today and actually drive through their codebase and show what customizations have been made, and give them an inventory of, say, 80 or 100 customizations that have been made in the software.

“Then we can ask the customer: Are those important? Are those relevant—still relevant today as they were 20 years ago? And the answer may be: a) ‘No, it’s no longer relevant’; or b) ‘We’re willing to change them, given where we’re headed.’

“There’s an additional point to this,” Hurd said. “It’s now a different world because of the new codebase. In some cases, our development organization is now so scaled, because the number today–as opposed to those old applications, they all had different platforms, they were all built in different environments—everything (Oracle does) is built in a similar environment, so the productivity of our developers is materially higher. So in many cases—some, but not all—the features that had been put through customization (in the past) have been turned into features of the core of the product in SaaS.

“That’s point one. We can actually 'extend' these applications—a key word, as opposed to ‘customizing’ these new applications. If there is something you just have to have, without affecting the core codebase, we can write the additional piece of code, connect it via API (application programming interface), and, if you will, ‘replicate’ the customization. But not replicate it by changing the code, but by extending the code. That’s very different.”

API or Application Economy?

That brings to mind this question: Is it going to be an API economy going forward, or an application economy? We at eWEEK have been discussing this for a long while. Oracle went out last year and acquired a company, Apiary, to handle API management for its customers.

“There will be both,” Hurd said. “You will see continued addition of features and logic, AI tools integrated into the core applications themselves—as opposed to the old on-premises world, where you get an upgrade every three or four years. Now you’ll get one every quarter, and you’re going to get two or three hundred new features every quarter. That said, there will never be everything everybody potentially will want, but the opportunity to have an app library that allows you a suite of microservices that you can now connect via API—that will be a material market.”

Part II, coming soon: Hurd talks about using Oracle’s SOAR to help companies move to the cloud, competitors and some other interesting items.

Chris Preimesberger

Chris J. Preimesberger

Chris J. Preimesberger is Editor-in-Chief of eWEEK and responsible for all the publication's coverage. In his 15 years and more than 4,000 articles at eWEEK, he has distinguished himself in reporting...