Janet Perna has seen IBMs data management business evolve from being a relatively small unit heavily weighted toward a DB2 relational database for the mainframe to being a billion-dollar software business with database, content management and integration products for mainframe and distributed computing environments. With an eye on perennial database competitor Oracle Corp. and the growing presence of Microsoft Corp. in the data management world, Perna, general manager of data management in IBMs software group, clearly isnt dwelling on the past. In an exclusive interview at her Somers, N.Y., office, Perna talked with eWEEK Department Editor John S. McCright and Senior Writer Brian Fonseca about IBMs On Demand computing initiative, the next version of the DB2 database (code-named Stinger), and her vision for the real-time information infrastructure.
What are your big goals for IBM data management in 2004?
Our overall macrogoal is to continue to grow and continue to gain market share. We have now topped Oracle in database market share, we are the market-share leader in enterprise content management again, according to Gartner [Inc.] and analyst reports. The objectives that I give my team—its really about being the best at providing customer value, which says best technology, the best solutions—either integrated solutions through IBM or through our partners—its the best customer support and service and best overall time-to-value for our customers. We believe if we can do all of those things, then we will continue to grow and we will continue to invest in the areas that are important to our customers.
We talk to C-level executives. Things that are on their minds are things like, “How do we improve operational efficiency of our business, how do we get better return on assets, how do we grow our top-line revenue?” Over the last few years, companies were focused on reducing cost, everybody was kind of bolted down and [in] cost containment. But now the economy is reported to be looking a little bit better and companies are again focused on, “How do we grow and acquire customers?”
How will you take advantage of the improving economy?
We really think as a company were uniquely positioned to be able to offer them an integrated set of capabilities that span from business consulting services to servers to storage to a middleware infrastructure for helping them build the next generation of applications. This next generation of applications will be built using service-oriented architectures, using integratable components and allowing them to move much more quickly than they have in the past.
The other thing that businesses are starting to think about is how their businesses and applications have evolved, theyve pretty much been stovepiped. Theyve been working on gaining operational efficiency within a particular stovepipe of activity, like manufacturing for example, [and] the thing that is new here and the area that theyre working on is the integration of these business processes across the enterprise. Thats about the integration of the CRM [customer relationship management] systems to [enterprise resource planning] systems to supply chain management systems to now really be able to provide operational efficiency and better return on assets.
So, you look at On Demand and what is that really about? Its about enabling companies and enterprises to respond dynamically to changes in their environments or the economy or industry, through this autonomic business model, which is this horizontal integration of capabilities. Thats IBMs overall strategy and what we are delivering for our clients.
How does DB2 and IBM data management fit into that?
You come underneath that, and information is a big part of anything. Without information there is no transactions, there is no IT, theres nothing to look at. As you look at these stovepipes out there today and whats been built the last 35 or 40 years, the information is really bolted today to whatever the business process was. Some of it is in relational databases, some of it is in file systems, some is in e-mail systems, some of it in imaging systems. There is very little organization of that information within a stovepipe, and then if you start thinking about across stovepipes, the integrated business view of the problem becomes even more challenging for our customers.
If they want an integrated view of their clients, we have a stovepipe business that was mortgage, and another was credit cards, another one was checking. Each one of those stovepipes of business units has its own information about the same clients. What [companies] want to be able to do as an enterprise is put together this information about clients, for example, and get a consolidated view of the clients. Why they want to do that is it enables them to better service the clients for sure but also do things like cross-sell across their client base.
Does that mean IBMs Content Manager gains importance because that is one place where you can take all sorts of data and manage it?
There for sure is information that is going to be stored in content repositories. There is already a boatload that is stored not only in IBM repositories, but in [repositories from] many other companies, and in e-mail systems from IBM and others. When you start looking at customer problems associated with doing this, you have a growing amount of information about clients. Second, [the information] is in all different formats and in all different places, its not all in IBM stores or in any one vendors stores, so thats a challenge.
The third [challenge] is compliance. There are over a hundred compliance regulations that companies will have to implement over the next 12 months. Not every company is going to implement 100, but if you go industry by industry, they each have their own compliance regulations. And most of these regulations have to do with information. HIPAA is about information. Sarbanes-Oxley is about supporting processes with information—this is how you say youre doing something, how youre really doing it, what is really the source of the information youre losing, and the traceability of it. Basel-II is really about traceability and accounting information.
Theres a tremendous focus now on information, quality of information, location of information, traceability of information that companies have to deal with over the next couple of years. So when you look at what were doing with information management within IBM, its really providing infrastructure to help companies deal with these challenges of massive amount of information, complying with regulations, and really being able to better leverage that information to drive operational efficiency and provide better customer service.
What areas are you investing in to do this?
Things were continuing to invest in—of course, DB2. DB2 is the centerpiece of our information management capability and offerings. DB2 for strong, not only structured, record-oriented information, but DB2 for storing other forms of content. That leads to [IBMs] Content Manager [for storing] images, documents, digital media and e-mail.
Then weve made a couple of acquisitions over the last couple of years. One is in electronic records management so we have the DB2 Records Manager, we have the DB2 Documents Manager, and … content management [that] has been branded a Lotus product, Lotus Workplace, where content is stored in a DB2 content repository. Were embedding analytics into the DB2 engine, things like data mining algorithms [and] OLAP capabilities into the database engine. The reason were doing that is to be able to provide real-time analytics of information.
How much easier does data mining become with the analytics embedded into DB2?
You still have to use a modeling tool to build the models. The thing that is important here, you have a class of applications weve talked about for years, which is OLTP. Now what were seeing happen is a new wave of applications around analytics. What companies want to do now is really leverage information. How do they optimize the end-to-end business process? It really is about starting with the customer … [going] back out through the supply chain, being able to determine in real time where are the bottlenecks in the process, what do you have to do to really optimize the process. In order to do that, you need to provide real-time information and real-time analytics over that connection as its occurring.
This is one area that has begun to be talked about as business process management. This whole area is about now the optimization across all these silos of information. Its much different than what weve talked about in the past and what weve delivered in the past. DB2 Information Integrator plays a key role in this vision [by providing] the ability to federate information, to replicate it, to federate a Web service or information coming though a Web service, the ability now once I have that information coming into a database to be able to do analytics in real time on it—real-time scoring of customers, predicative analysis based on transactions coming in—things like that are important to have and support within this infrastructure.
Oracle has applications and want more. Microsoft has applications. Even Sybase has mobile applications. Why isnt IBM getting into the applications game?
Our decision was made a few years ago that we were going to exit the application business. One of the big reasons we did that was you begin to alienate the application partners, and weve seen that through Oracles actions [and] Microsoft has alienated application companies. We really believe that if you look at IBMs core competency, where our strengths are and what our clients see as the value of IBM, it really is delivering this infrastructure of storage, servers, middleware and consulting services and partnering with the application providers that really provide them a level of choice on the application level.
Who are you trying to address with Stinger? What is its timeframe?
If you look at any new version of DB2 there are design objectives, One is continuing to make life easier for [database administrators]. It really is around reducing total cost of ownership, that is around autonomic capabilities, high availability, performance, and ease of application development.
In the industry, when you talk about business intelligence, they think of tools [from vendors] like Cognos, Business Objects and MicroStrategy. Im thinking about a broad set of analytical applications and analytics that are going to be embedded in every application. You see it already. You see Siebel analytics in a Siebel environment. You have the same with PeopleSoft, SAP. Every application in order to do this real-time optimization of the business process will have to have analytics embedded in the application. Its not going to be this tool sitting out there. Its going to be having a platform that allows you to do that real-time analytical processes.
So the thing were building is that infrastructure and reference architecture for this analytical platform, which is really built on DB2 as the underlying data store. Information Integrator, partner capabilities are part of that, and cube views.
When will we see enterprises managing structured and unstructured data in the same database?
It will eventually come together, but it will not be in the next two or three years. Where IBM is uniquely advantaged here, if you look at FileNet they dont have a relational database. Oracle doesnt have the content stores so they have no way of storing very large collections.
Security has become a very high-profile concern for enterprises. What is your data management group doing to answer that?
Security has always been a huge focus. With DB2, you look at the roots of DB2 on the mainframe, and it really is probably the gold standard for highly reliable, secure, high-performance technology.
What do you make of Oracles positioning of its 10g database as the best option for grid computing?
10g, we know what it is. Some of the things Ive read, Oracle users are saying, “What is the big deal with 10g?” With high-level information I have about it, its like RAC [the Real Application Clusters technology in the Oracle 9i database]. RAC hasnt taken off in a huge way from what I can tell. The installations out there are pretty simple failovers. You look at this [industry], and it really is no longer about the database. Oracles been pigeonholed as this client/server database/Unix platform.
The database continues to be important to us, but what we see evolving here is this whole information infrastructure that is really what companies are going to be [building on] in the future. The ability here to be able to leverage and integrate all forms of information. Its not just the information thats in a relational database, its being able to come in and search and pull information whether its in a content store, whether its on the Web, and be able to integrate that information and do it in real time.
Thats where we see the real value of our clients moving.
Where is IBM at on the database pricing wars?
Weve got a lot of clients whove called us in because of Oracles predatory pricing, thats how we got in the door. They said, “Im really ticked off at Oracle,” took a closer look at DB2 and said, “I want to go that way.” Were the ones who really put a world of hurt on them by providing their clients with a better value proposition.
Be sure to add our eWEEK.com database news feed to your RSS newsreader or My Yahoo page: