Sybase is regaining ground in the database market. Propelled by last summer's launch of Version 15 of Adaptive Server Enterprise (ASE), Sybase's database license revenue for the second quarter grew 22 percent over the prior year -- the fastest rate in the past 12 years, according to CEO John Chen. But there's more: At a time when many enterprise software vendors are looking to reinvent themselves as Salesforce.com clones, Chen wants to turn Sybase into an old-school channels powerhouse, a la Microsoft.
In an interview at Sybase's annual TechWave conference this week in Las Vegas, Chen expounded on that plan, as well as why the vendor's push into data integration doesn't threaten Informatica or IBM's Ascential lineup -- at least for now.
Excerpts from the interview follow:
In the second quarter, 46 percent of your revenue came through indirect sales. Will indirect sales make up the majority soon?
When I started, 20 percent of our sales were indirect. I would love to be 75 percent indirect, 25 percent direct, though it's not our stated goal inside the company. Because what Sybase is really good at is building technologies, creating [intellectual property]. What we don't do as well is go into hand-to-hand combat with the big boys, who have more salespeople than my entire workforce.
We do want to keep our direct enterprise customers happy. To do so, we want to stay like a limo driver, not a taxi driver. Meaning, we don't meter you for everything. Oracle, they charge you for everything they do. Maybe they can get away with it. But we train our people not to say, "What's changed?" when a customer has a problem, as if it were automatically the customer's fault, even though nine times out of 10, they did change something.
But even Microsoft is getting a lot of pressure to explain how it will fit into an industry that is supposedly moving toward a software-as-a-service model. If that's happening, why go back to a channels model?
The interesting thing about running a business is that nothing is absolute. Given what I think Sybase excels at, we can improve our margins and expand our reach by expanding channels today. And next year. And the year after that. Everything you mentioned about SaaS -- there's no money in it now.
That's not to say we should fight a subscription-based model. The trick is when to get into it. We just bought a German company, Solande. Their business is done completely through Internet, aimed mostly at [small and midsize enterprises]. One of these days, Solande will become a full enterprise product. But it will take a while, intentionally, because we don't want to destroy our relationships with Informatica or Ascential.
So we like Solande's hosted model, though it's so small it's really just a dabble for us. We are thinking of creating solutions which some telcos and others would want us to host. We have a lot of ideas kicking around that would expand our market and not intrude in our indirect efforts.
What about open-source?
Our big customers -- banks, governments, health care organizations -- are very unlikely to use open-source during my professional lifetime for mission-critical needs. When you're throwing trillions of dollars around the globe, you're not going to think, 'Let's save a little money by going to open-source.' That's insanity. Though it is different for departmental solutions, or internal systems like payroll.
We haven't bought open-source products like IBM. IBM's strategy is an interesting one. It's not a bad one, but only they can afford to do it. They intentionally create in some ways their own competition. If I give you the choice of two, and I own both, then it doesn't matter what you pick. It's a good strategy, though it will cost you a lot of money along the way. But what's a little money to IBM?
What Sybase does is different. We develop code around a certain open-source standard like Eclipse. Reason?
All these people like IBM and BEA are committed to Eclipse. We can develop a developer base. It's like seeding the market. My technology later on is compliant with the community. It also cuts development space. I really believe in that strategy.
Who is buying ASE 15, which you released at this show last year?
Three quarters are existing customers who bought because of features like partitioning and encryption, which we think is the best in the industry because it's so fresh. We will have clustering available in the first half of next year.
One quarter of ASE 15 customers are new. They are competitive wins, many of them on Linux. We won a lot of Microsoft NT users. Our biggest partner on this is IBM, with their P series of servers.
With Web 2.0 start-ups on the rise, are you losing key employees, as you did during the late 1990s?
No. This is not a challenge to come and recruit our people, but we have very low turnover: 10 percent to 12 percent per year, which is low in the industry. It's not as fashionable to jump out of a bigger company into a start-up, because everyone still remembers the nightmare of 2001-2002. Though I must say we're doing so good in mobile, my mobile guys do get more interesting offers.
On mobile devices, is the form factor slowing down service rollouts and software innovation?
Yes, the user interface is still not the best it can be. However, let me say this: My 16-year-old daughter, she can type text messages with one hand. She can be eating with one hand and typing with the other. And she's not unique. We may have gotten used to a certain form factor, but the younger generation isn't as bothered.
I think voice input will be big. Orange Telecom in France is using our voice-to-text technology for directory lookups. And we will likely use it with the Beijing Olympics. It won't just be Chinese -- it will be in six languages. I can't go into much detail except to say, imagine how amazing it'll be when you get voice and GPS working together.
Have you had customers lose devices, like the U.S. Veterans Administration?
Yes. But we wouldn't know if any data was compromised. We have customers who have used our software to invalidate and lock out a stolen device over the air. Obviously, everything is also password-protected and encrypted. There's enough precautions, but it's still not foolproof. We can't do the Mission Impossible thing, where the device starts smoking and destroys itself. We need to get Intel to invent some way to either release a corrosive on demand, or apply maximum power to the chip and let it overheat.