PeopleSoft focuses on business process integration

CTO Rick Bergquist discusses the impact of Web services in moving information in real time

020709hnbergquist.gif
AS CTO OF PeopleSoft, Rick Bergquist has seen a lot of technology come and go. But with the advent of Web services, Bergquist sees a tremendous opportunity to reinvent business as we know it today. In an interview with InfoWorld Editor in Chief Michael Vizard, Bergquist says the time has come to focus less on the enabling technologies around Web services and more on the impact they will have on business processes.

InfoWorld: With all the focus on Web services these days, what impact do you think this technology will have on enterprise applications?

Bergquist: Our take on this is that Web services are an enabling layer for us. [They’re] enabling revised business processes and the streamlining of business processes. We’re referring [to the] trend as “real-time enterprise,” which is a key concept because the data you are using in your organization is actually old. [But] if you’re using Web services, you can now get to things. You can get information to your customers, your partners, your suppliers, or whatever the case is. And the good news is we’ve finally got a technology that both the Microsoft camp and the non-Microsoft camp are agreeing to. The big thing that happened five years ago was the browser enabled any user to get to any system. The impact of Web services should really be [that] any system can get to any system.

InfoWorld: For years, people focused on application integration. Now we are talking about business process integration. What’s the difference?

Bergquist: Business process integration carries a broader connotation. If you look at it, your business processes normally extend beyond the boundaries of the organization. Usually, application integration ended up being about getting a human resources system talking to a financial system and the focus was all internal. To me it’s a difference of scope.

InfoWorld: So where will people use Web services first?

Bergquist: If you’re looking at Web services, [they] will probably be first used to do application integration. XML is certainly more flexible than anything that came before it, and you’ve got to start there.

InfoWorld: What impact do Web services have on the debate over best-of-breed applications vs. suites from a single vendor?

Bergquist: Best-of-breed becomes more feasible. In a commercial world, 80 percent of our sales are [from] one or two product lines as opposed to all four that we carry. So best-of-breed exists today and people are dealing with large organizations that do buy on a best-of-breed-type of basis. We can point [to] hundreds of customers that have a mixture of us and SAP. Web services make best-of-breed more cost-effective. But I don’t think it’s going to go as far as the extreme of componentizing everything. What you’re looking for from a vendor is you want software that is pre-built, pre-integrated, tested, and you want someone else to incur the pain of integrating all these components. So what you want to have is the ability to have fairly large, modular chunks that have the functionality so you can minimize the number of upgrades you have to do, but let the vendor do it.

InfoWorld: How has the current economy changed the nature of the competition in the enterprise application space?

Bergquist: On one hand, there’s always an appeal to go with the suite because you have the synergy of common data models, common thinking about business processes, and so forth. But what we’re finding in today’s economy is that people aren’t doing a big-bang approach where they buy the whole suite. The reality is that the big high-risk, high-reward business scenario went out five years ago because basically there were too many failures. Customers may not mind going with a single vendor, but they are going to do it a piece at a time. People are really looking for low risk/high rewards, and they’re looking [to bite] it off piece by piece. And all of these expenditures are getting really high scrutiny from the CIO and the CEO, who all want [to see] the return in the same year that they buy the product.

InfoWorld: Speaking of failures, CRM projects that failed seem to be at the center of the storm lately. What went wrong around CRM?

Bergquist: Part of the problem with CRM is it didn’t consider the whole business process. The problem with CRM is that it was focused on a silo of information and it didn’t transcend to the overall business process. And there weren’t clean ways of dealing with that so the business processes [were] not complete. The other issue is that most of these systems started out as sales force automation systems that were bought with the intent of controlling the sales force. I don’t think you control a sales force. What you should be doing is enabling a sales force, because what you’ll find is that the sales force doesn’t feel that they want to use the system. What sales people want from an SFA system is something that passes them leads and then grudgingly — if it’s easy — they fill out the return, which is a sales forecast. The problem is the systems weren’t usable. For most mobile applications, it took forever to synchronize them. It was just unmanageable and it’s using the wrong technology.

InfoWorld: What’s your approach to CRM then?

Bergquist: The total footprint we’ve got is less than 2MB of data. I have a Rip Van Winkle test, where a salesperson goes to sleep for a year and then wants to synchronize everything. When we synchronized a year’s worth of changes, on a laptop it took six minutes, on a PDA it took 12 minutes. So instead of bringing down every change, we just said, “Isn’t there a good copy somewhere?” “Yeah, it’s on the server.” “Why don’t we just go grab it?” So we brought it down. You get the net change of what it is and then you radically reduce the amount that has to be synchronized; you don’t have all this processing. You’re replicating the data down, but you’re not replicating every database change. The other cool thing is that it’s a Web service so it connects to the rest of the business process. So we re-thought CRM and why we think it’s a much better way of doing it.

InfoWorld: Another area that PeopleSoft is moving into is the corporate portal space. But given the number of offerings that already exist, why bother?

Bergquist: We looked at the marketplace and [decided] we could either build it or we could, in essence, OEM someone else’s. We chose to build our own because we can handle the volumes better than anything else based on our technology. And we handle some of the more complex things like multilingual aspects of our applications. At the time we made that decision, none of the portals were multilingual, so we chose to, in essence, use our technology to build out portal technology.

InfoWorld: What’s your take on business intelligence as it relates to your applications?

Bergquist: You’ve got to bring together the analytics so you’re harvesting content. The problem that we’ve been seeing is that in the past analytics were created as a separate thing. Our view is analytics can’t be after the fact. It needs to be real-time, so what you need to do is to build out the analytics but then embed it into the business process. We bring the content into the enterprise warehouse and then we put in the analytics as an embedded part of the business processes.

InfoWorld: Everyone in the enterprise applications space throws around the phrase Web-based applications. In a PeopleSoft context, what does that mean?

Bergquist: We push PeopleSoft 8 as a pure Internet architecture on two levels. One is the user interface. For us the user interface should have no code on the client, it’s basically HTML only. The key there is that it gives you universal deployment. At the interface level, it’s Web services. It’s UDDI [Universal Description, Discovery, and Integration], it’s SOAP [Simple Object Access Protocol], it’s WSDL [Web Services Description Language]. That to me is what it takes, both levels — a user-to-system and system-to-system architecture using everything in the Web. If you look at Oracle, they’ve got a Java client. We think that’s wrong. The real problem with client/server was the high cost of maintaining the applications done on the client. The way to reduce that cost is to get rid of it, not to minimize it. SAP is basically saying they were going to convert. To date, they have a browser front end into the R3 applications. What SAP has are client/server applications on a browser.

InfoWorld: What’s your take on the viability of ASPs?

Bergquist: The ASP wants to turn software into a factory. They want, in essence, to reduce their costs by repeating a process over and over. Our advice is to minimize customizations for any customer, but at the end of the day there’s a strategic differentiation between you and your competitor and everyone’s looking for [a strategic difference] so you end up customizing the software. If you think about people, 3 percent of the DNA is different across people and you get a wide variety of changes. As soon as you start having things that are unique, in essence the factory model for repeatable process seems to break down.

InfoWorld: So what is your best advice for customers today?

Bergquist: There’s a lot of hype [about] Web services standards. They’re just getting the gauges of the track lined up. The real interesting thing will be what’s going across Web services. What is the information they’re exchanging? Hang on for that and look at how you’re going to change your business when we can basically connect you with anything. The most exciting part of Web services, it’s not the standards, it’s how it’s going to change the business process and move information in real time.

Source: www.infoworld.com