The Data Game
An Interview with Peter Auditore, President of Syndicated Services
“Data. Anytime, anywhere. That’s our vision,” says Peter Auditore, president of Syndicated Services, and head of the Business Intelligence/Data Warehousing (BI/DW) research program for Survey.com (Mountain View, Calif.), an Internet-based primary market research initiative. With clients such as IBM, Oracle, Microsoft and Sybase, it’s his business to know data. And the company recently completed a series of studies specifically aimed at understanding user requirements in the BI/DW marketplace.
“Over the past seven months, we’ve done more than 5,000 surveys online regarding BI/DW, exploring such topics as Customer Relationship Management (CRM), data warehousing of ERP information, data warehousing in general, by hardware, services vendor, in-house spending, and so on,” he explains.
So, when MIDRANGE Systems went looking for an insider’s perspective on business intelligence and data warehousing issues, we went to Peter. In the following interview, he sheds light on the subject both from the standpoint of a consultant, as well as that of a user, with a candidness and enthusiasm that gives data analysis a whole new twist.
MRS: As someone who specializes in business intelligence, how do you define BI?
PA: Business Intelligence is an interesting term—does it mean going through a dumpster and picking out competitive information? To some people it does. But in my mind, it is a strategic initiative within an organization to leverage legacy and real-time data in a way that empowers people to make better decisions overall.
MRS: Findings from one of your most recent studies indicate that business intelligence is one of the fastest-growing market sectors in the IT industry. Your figures show an average annual growth rate of 43 percent worldwide, expected to reach $148 billion by the year 2003. What do you believe is driving this growth?
PA: If you look at BI overall, one of the things that’s driving it right now is e-commerce, especially B2B e-commerce. Let’s say I’m a supplier of automobiles and I want all my dealers to know what the inventory is. If I take my ERP information and I data warehouse it, then I can provide it real-time 24x7 for B2B e-commerce for all my dealers. Knowing and understanding the customer is a key ingredient in this effort. That’s where another BI driver, CRM (Customer Relationship Management), fits in. If people don’t feel like they’re doing CRM or incorporating CRM-oriented applications in their company, they’re afraid that they’re not going to be able to maintain their competitiveness.
|"In the end, the real beauty of business intelligence/data warehousing technology is that it brings me closer to my customers."|
What’s interesting is that if you survey people in different market segments, they get confused about what CRM actually means. In retail banking, for example, when CRM is mentioned, they don’t think it applies to them. They don’t want a relationship with their customers—they want to know who the best ones are, who the worst ones are and how to get rid of the ones they don’t want. In fact, I interviewed Wells Fargo recently, and that’s exactly what they told me. Usually when I ask a company, “What does CRM mean to you?” it means customer profile analysis, or knowing their customers. That’s why I like to call this Customer Knowledge Management, or CKM. But whatever you call it, companies are beginning to realize that they have to have a one-on-one relationship with their customer base. So that’s another major driver in the BI market.
Thirdly, a lot of companies also want to provide access to human resources information from ERP systems, and that’s driving the use of BI tools, too.
MRS: So, what has really accelerated all of this to the level that Business Intelligence is now the new buzzword?
PA: Well, actually, BI isn’t a new buzzword. Business Intelligence is a term that was coined a number of years ago by Howard Dresner, an analyst with GartnerGroup. What has happened recently is that the market has reached an early mainstream level—in that the technologies have been tried—people know that they work, and now they want to move the technology out to other areas of their business to improve their business overall.
So some of the things that are really driving this are: 1) the ability to do CRM and increase your revenue and sell more to your existing customers; and 2) Web-enabling of the data warehouse to reach a larger audience of people. The B2B stuff is driving a lot of data warehousing on the front end and will continue to drive it for the next three to five years.
Web-enablement of the data warehouse is key. It’s one of the reasons that a company called Business Objects (San Jose, Calif.) is beginning to take off right now. They have effectively provided a scaleable, Web-enabled front-end tool for query reporting into data warehouses. With this kind of capability the data itself becomes more important. Look at Survey.com—we’re the only market research company that guess what, gives you data. I’ve got a guy selling Excel files to companies for $5,000 each.
MRS: So that they can then go ahead and apply their own BI tools?
PA: Exactly. So they can do their own processing, their own analytics—whatever they want to do.
MRS: In general, how good are people at taking advantage of BI tools today? You’re saying that you’re selling your data, which is a testimony to the value of these kinds of tools. But to what extent have users really adopted this technology? Are users taking advantage of multi-dimensional databases, for example?
PA: This technology is being used globally very effectively by many organizations. Even police departments are beginning to use the technology. Over the last three to five years, the tools have improved a lot. The front-end tools in particular have become much more user friendly—a semantic layer, for example, protects users from SQL programming so that they can do online analytical processing (OLAP) without really knowing SQL calls. And, all different kinds of query reporting are possible.
Let me give you an example. I was in London two years ago and went to a police station where they had data warehouses of parolee information, and they were predicting when the parolee would commit his next crime. That’s the extent to which this technology is being used. It’s not just for boring sales analysis by region, for a chocolate company, or whatever. It’s being used to profile and model the customer base. In fact, even the World Wrestling Federation uses data warehousing and business intelligence initiatives to understand its customers better—who comes more often, who to target market, who to sell more things to, and so on.
As to whether users are fully exploiting the tools, that’s another story. Generally speaking, for example, the use of online analytical processing tools (OLAP) is overstated. The majority of people in any organization are NOT using OLAP. In fact, our findings indicate that, on average, less than 20 percent of those using BI tools actually use OLAP. Most BI users employ query and reporting tools and/or simply read reports in BI repositories that are archived or sent to them.
MRS: What are some of the issues and challenges that companies’ face in implementing BI/DW tools?
PA: One of the most difficult challenges right now for many organizations is Web-enabling the data warehouse—creating thin-client front ends to data warehouses. And that’s simply because a lot of the front-end tools have just begun to evolve. Even leading vendors like Cognos (Ottawa) haven’t done a good job of Web-enabling their tools. That’s one of the reasons that these Business Object guys are taking off right now. They have a product called Web Intelligence to address this.
Also, integrating all of these different kinds of data is a big issue. That’s probably where 40 percent of all revenue is spent—in just bringing in different kinds of data on a daily or weekly basis. That’s a major issue, because data comes in from many different sources and in many different formats.
For example, for CRM you will probably want to do some sort of legacy sales data integration, then integrate some call center data. And all that stuff has got to be cleansed and modeled and profiled so that you know that it is accurate and that it represents what you’re looking for. Long before you actually move data, you have to model your process and your project and be clear as to what you really want to know about your customers. And that in itself is no small task. But, according to our research, the real challenge is actually moving the data. I mean we still have a lot of people doing custom COBOL programming out there to accomplish this.
MRS: Do you think that’s going to change?
PA: No. I think it’s going to continue to be a major issue until there’s some sort of standard database format around the world, which I doubt will ever happen. I think there’s going to be all kinds of disparate and heterogeneous data out there for a long, long time to come, and that’s going to present some opportunities for a lot of vendors.
Basically, to do a data warehouse, customers are telling us that what’s most important and where they spend the most money are in these two key areas: Web-enabling a data warehouse, and then getting data in and out of it that’s reliable.
MRS: So, at this point, who are the major players in this space? Are there any new players on the horizon?
PA: As I said, on the front-end side, Business Objects is a major mover right now—Cognos fits in here, too. Microstrategy (Vienna, Va.) is another company on a roll, because of its use of OLAP and reporting tools to deliver an end-to-end BI solution. But IBM has the largest market share—not only because of its technology, but because of its Global Services division, which is very, very strong. IBM mainframes, AS/400s and Unix-based systems all play a role in data warehousing. But since one of the greatest challenges centers around integrating all this disparate information and data, and no one company has a single product that meets all the needs of all the companies, it’s important to be able to provide that integration expertise. That’s where IBM has an advantage right now, because of its Global Services division.
Basically, the two key players in this whole data warehousing space are IBM and Oracle (Redwood Shores, Calif.). These guys dominate in market share, mind share and revenue. In the future, however, Microsoft is going to play a much larger role in data warehousing, because it will make data warehousing a lot more affordable. I expect we’ll see data warehouses or data marts rolled out on a SQL server platform that has integrated OLAP services built in to it. (It doesn’t hurt that Microsoft also offers a front-end called Excel, which is pretty ubiquitous in terms of analytical tools.)
You can also expect to see a lot of added functionality like pivot table services added to Excel—OLAP basically for free. And in the future, I think Microsoft will significantly impact this market by reducing the overall cost of data warehousing and making it available to many, many, many more people, and many, many more organizations. Today, most data warehousing projects cost about $1 million to $3 million on the low end and can cost upwards of $10 million to $20 million. Microsoft promises to change all that. Imagine SQL Server on inexpensive Intel platforms with Excel already on the desktop! It’s going to be a whole different game down the road.
MRS: Overall, what can IT managers and decision makers who rely on BI/DW tools expect to see happening in the next few years?
PA: Well, storage is one area that is going to receive much more attention in the years ahead—it’s a very important part of BI. In fact, we see the storage market growing significantly in the future, and we see data warehousing expanding globally and being put to greater use in all different kinds of companies.
Generally speaking, I see business intelligence becoming ubiquitous throughout the world. It will be very inexpensive and all kinds of companies will begin using this kind of stuff, not just large Fortune 500 or Fortune 1,000 companies.
MRS: Where does the AS/400 fit in this picture, in your view?
PA: The AS/400 is definitely growing as an e-commerce server. But our findings indicate flat growth in the data warehouse area for the AS/400. The hottest data warehousing platform currently is the Sun Starfire—most of the data warehousing is moving off of the mainframe to Unix systems.
MRS: As a user of BI tools yourself, what kinds of thoughts might you offer to others? What is it that’s most exciting to you about using these tools?
PA: What’s most exciting to me is that I can leverage data in many different ways. The smarter it makes me, the closer I get to my customers. And this is all about customers. If you really look at what’s going on in the market, there are so many companies, there are so many advertisements. Do you remember the old television show, “Cheers?” “Everybody wants to go where someone knows your name.” I love that.
When someone sends me an e-mail and follows up with me if I bought something from the company, I think that’s great—that’s personalization. In our company, I use data warehousing technology not only to facilitate data analysis and provide my customers with better services, I also use it to know my customers and to maintain and leverage our survey respondents. Last year, we had a million people come to our Web site and take surveys in a variety of different areas. And every one of those people had to be entered into a database, they had to be managed, and they had to be profiled. In the end, the real beauty of business intelligence/data warehousing technology is that it brings me closer to my customers.
Peter J. Auditore is the president of Syndicated Research and director of the BI/DW research program for Survey.com. He can be reached at mailto:email@example.com.
Related Editorial:Hummingbird Unveils Metadata StrategyCRM and Data WarehousingBusiness Intelligence a Major Post-Y2K Thrust, Surveys Confirm
Related Information:survey.com (new window)