Mark Hurd is standing. It's his habit to do so, but it also eases a little back pain, the result of a bit too much tennis. Otherwise, the Oracle president seems comfortable, at ease, much different from how he appeared in his occasional outings while running Hewlett-Packard, where he seemed beleaguered and boxed in, but also where he infused the company with a business discipline that boosted profits and made Wall Street happier.
If Oracle founder and CEO Larry Ellison's fangs are scarier than those of HP's board, Hurd doesn't show the marks. In a recent interview with me, he asked for the freedom to just talk, neither on the record or off--we'll sort it all out later. It's as if Ellison's unapologetic brashness allows his two top lieutenants--Hurd and Oracle's other president, Safra Catz--to speak comfortably and freely.
On the occasions I've had to observe Catz in action, she hurls barbs like it's a Charlie Sheen roast. In a recent talk this year, Catz was playing up how easy it is to implement and use Oracle's Fusion application suite, contrasting it with the main competition. "In SAP they will be like ctl-alt-F12," she said, adding with a smirk: "I exaggerate. My German isn't that good." She takes no prisoners.
Nor does Hurd. When I quizzed him about SAP's cloud offerings, he shot back incredulously: "You mean Business ByDesign? Come on." And then he gave a soliloquy on building enterprise software.
[ What are IT innovators doing now with big data? See 10 Lessons Learned By Big Data Pioneers. ]
Hurd's got a right to be a little cocky. Oracle has had another succession of stellar quarterly results (more on that later), and many key aspects of its business, including new license revenue for applications, databases, and middleware, along with support revenue, show strength. During its latest earnings call on Tuesday (Q1 2012), Hurd emphasized that amid this strong growth, practically nothing stood out--all industry sectors, all geographic regions, all product lines performed well.
The company is bullish on its hardware prospects as well, claiming that while it has taken a while to clean up the business aspects of Sun, hardware growth is coming soon. The Sun acquisition narrowed Oracle's comfortable operating margin for the past year, and while Oracle hardware sales haven't lived up to expectations (they were down 5% in the most recent quarter), Catz pinned that result on Oracle's insistence on selling hardware more profitably, and turning away deals that weren't.
On Tuesday's earnings recap with analysts, Ellison made this aspect of Oracle's business crystal clear: "I don't care if our commodity x86 business goes to zero. We don't make any money selling that...we are phasing out that business." Oracle's focus, he said, is on selling its hardware systems based on its own intellectual property--what the company calls "engineered systems." Ellison also said that overall top line hardware sales will show growth in Oracle's next fiscal year. In other words, the company is just fine with lower hardware sales until it can get its profit margins in a better place.
Catz said that hardware gross margins climbed to 54% in Oracle's first quarter (from 48% last year), and Hurd said Sparc sales grew. At this year's Oracle Open World, the company will unveil a new Sun Sparc Supercluster, running the company's new T4 Sparc processor, increasing performance by a factor of five, according to Ellison.
The company's goal, Ellison said, is to get overall operating margins back to pre-Sun levels--around 46.2% for FY09 (vs. 42% for the most recent quarter, up from 39% a year ago.)
Oracle has gotten stronger while reshaping almost every aspect of its business, save for its core database, which has dominated for a long time. But even then, Ellison schooled analysts on Tuesday about how the company's flagship product has evolved from its roots as a relational database to one that can handle all data types, including unstructured data, making a point to call out Autonomy, the content management company that rival HP has targeted as its transformative acquisition--one that Ellison said Oracle passed on because the price was "absurdly high."
Although it seemed as if Gaudi's Sagrada Familia would be finished before Oracle's Fusion application suite, it's finally here (the software, not the Barcelona cathedral.) Its Exadata and Exalogic appliances are growing faster than any other Oracle product, the company says.
In some ways, then, this is a new Oracle.
All of the building blocks were in place before Hurd moved up the peninsula to Redwood Shores, but he is entrenched, as if he'd been at Oracle all along. Ellison, Hurd says, is the chief software engineer; Hurd's job is to bring that software to market. And he's interested in clearing up a few things.
1. Consumerization Helps Oracle
There are almost a billion phones in the world, Hurd says, but he predicts we're headed toward 3 billion. These are computers, and their users want and expect information constantly.
Hurd cites statistics that predict an 80% increase in urbanization by 2030. "You've got a Beijing now being built every other month," he says. "By 2045 the E7 will be bigger than the G7."
Hurd's generation might think talking to a call center to get information for, say, an inbound flight is adequately modern, he says, and it beats parking and walking into the airport and looking at the flight arrival screen. But the mobile generation demands to know where the flight is in transit, what its gate number will be. With "every human on the planet asking questions about all the data on the planet 24/7," data needs are growing, he says.
Oracle's databases house 65% to 70% of the world's data, Hurd says. Big Data? "It's another way of saying people want information," he says.
2. It's Not All About "The Stack"
The media portray Oracle as thinking just about vertical integration, Hurd says, that Oracle wants to own the entire stack, from the hardware to the application. What Hurd didn't mention is that's because it's often Oracle's refrain. Last year, Hurd's predecessor, Charles Phillips, now CEO of Infor, told InformationWeek that customers would save time and money by going with Oracle's stack.
Hurd doesn't retract those words. "Don't get me wrong. Stack is still good," he says. "But we want to be best of breed at every layer of the architecture." From the hardware to the operating system to the database to the middleware to the horizontal and vertical applications. Hurd emphasizes that Oracle sees a heterogeneous world, and Oracle's modularity fits that world well, he says. Oracle has organized its R&D around that thinking.
During Oracle's previous earnings call (Q4 2011), this focus on winning at every layer of the stack was also pointedly evident. A few excerpts from Catz, Ellison, and Hurd:
-- "Our software business is now even bigger than IBM's software business."
--"Our North America applications business passed SAP a while ago, and now, if you exclude Germany, we believe we have essentially caught up to them in Europe." (A Barry Bonds-sized asterisk, given that Germany is Europe's biggest economy, but the scholared business professor Catz knows that the day Oracle beats SAP in its own backyard is perhaps the day Catz's German becomes fluent.)
--"Sun's hardware system revenues were $1.2 billion for the quarter. We're running the Sun business on a more profit-aware model."
--"Our database business grew 28% in Q4 and 26% for the year. That growth rate is twice as fast as any year this past decade. We're already more than twice as big as IBM, the former No. 1 and current No. 2 player in the database business, and we're consistently growing faster than IBM and taking market share away from them."
--"It's the first quarter we've ever sold more than $1 billion in apps."
--On Exadata: "Our sequential growth was more than 50%. That's big growth, and it was broad-based. We sold to more than 150 different customers with close to half buying multiple systems. P&G, JPMorgan Chase, Apple, Fidelity, Boeing have helped us now cross the 1,000 systems mark in sold Exadata systems. We have 300,000 database customers running Oracle database workloads. Every one of them is a prospect for Exadata."
"The exciting news is, in addition to this, the Exalogic ramp has been even better than Exadata, and every one of our 100,000-plus middleware customers is a prime prospect for Exalogic."
Although Hurd says Fusion is coming out modularly, and that its human capital management app has been popular (Ellison claimed on Tuesday that Fusion has about 200 customers so far, and has beaten SAP for some deals), he spent a great deal of his time talking about Exadata. This, he says, is the best selling Oracle product right now. He even ventured that it's the industry's hottest product next to the iPad.
Hurd provided the example of Oracle's focus on the telecom sector, where it manages vast amounts of customer billing environments, across many carriers. He noted that one particularly large customer data set even runs Exadata on HP hardware.
Oracle is very much focused on building products for six industries--financial services, communications, retail, health sciences, utilities, and the public sector. All continue to perform well, and although financial services and the public sector are particularly strapped by economic factors, Oracle isn't providing any specific downward financial performance guidance to investors, according to statements Catz made during Tuesday's earnings call.
But Oracle does have its work cut out. Most of the world's biggest data sets are managed by the likes of data warehousing specialists Teradata, IBM Netezza, and EMC Greenplum. In a recent in-depth report on big data, we noted a recent Data Warehouse Institute study that indicates 46% of data warehouse customers say they will switch products in 2012. So the opportunity for Oracle, and others, is clearly there. (You can read our in-depth report for more on high-end data appliances like Exadata.)
Hurd says he thinks Exadata can win because most data sits in Oracle databases, and because vertical integration, this idea of building it on an all-Oracle stack, will become the norm, bringing management and performance gains and even cost savings.
Ellison said in Oracle's previous quarterly earnings call that the company plans to triple Exadata's installed base of 1,000 this year: "We think that's possible because Oracle apps run unchanged on Exadata," he said. "That's very different from IBM Netezza, which does not run Oracle apps or IBM DB2 apps or anything else, except applications that are custom-built for Netezza."
But Oracle's competitors claim that Exadata, which supports transaction processing and data warehousing, is mostly handling transactions. (An Oracle spokeswoman says that this is a misconception, but the company doesn't break out these numbers.)
Business intelligence and analytics applications, everything from Business Objects to Cognos to Oracle Business Intelligence Enterprise Edition, run on the data warehouses from the likes of Teradata and Netezza, and they're customer-specific and custom built for those platforms. I asked Oracle whether its data warehouse customers run custom-built applications, guessing that those also won't run on IBM Netezza or Teradata data warehouse appliances, but Oracle hasn't responded.
Oracle has hinted that it will address high-performance opportunities at Oracle Open World in early October. During the Q4 2011 earnings call, Hurd said Oracle is planning to add appliances, including "a large memory addition to Exadata...we've been the leader of in-memory database technology ever since we bought TimesTen. And that's both for transactions and for preprocessing."
Companies such as Quliktech, SAP, Tibco, and Spotfire are actually the ones writing the compelling in-memory story for now. But if Oracle can leverage the TimesTen in-memory database, which seems to have languished in obscurity, it may be an exciting response. Ellison also alluded to this appliance during Tuesday's earnings call, saying it would run "Oracle's database faster and more cost effectively than anything from IBM."
Wednesday, Oracle rolled out the Oracle Database Appliance, which InformationWeek editor Doug Henschen says you can think of as Baby Exadata. "This appliance gives midsize firms and departments a $50,000 platform for scaling up transactional or analytic workloads," Henschen notes.