According to Bill Baker, general manager of Microsoft's data warehouse product unit, business intelligence (BI) solutions help organizations understand customer buying patterns, identify sales and profit growth opportunities, and improve overall decision making. "What amazes me is that many organizations fail to empower their knowledge workers. I've been to organizations that tell me that 15,000-25,000 of their employees are decision- makers, but only 250 of them have decision support software. That's a staggering statistic," states Baker.

Baker joined Microsoft in 1996 and served for two years as manager of the Internet application server group with responsibility for the Internet Information Server, Active Server Pages, ODBC and OLE DB and the Component Object Model (COM) including the Microsoft Transaction Server before moving to the data warehouse product unit. Baker emphasizes that Microsoft is making data warehousing and business intelligence easier and more accessible by simplifying and integrating services into the platform itself ­ Microsoft SQL Server and Microsoft Office ­ and providing open interfaces to access and share data in a heterogeneous environment.

"As general manager, I oversee the entire product development and delivery of the business intelligence/data warehousing components of SQL Server. The components we build are fully integrated in the SQL Server package. We conceptualize what the product or next release of the product will be, write the specifications and do all the coding and testing of the product. We also make a lot of customer visits and review the support logs to determine what people find easy or difficult to use. We're very sensitive to the customers who already have our products and the questions they ask, because that indicates how well we hit the mark. What drives our whole team is the commitment to ensuring that our products work for everyone. That's a very complicated proposition and requires a lot of innovation on our part. It's very challenging, but also very energizing," states Baker.

"In terms of components, we have OLAP Services which we have rechristened Analysis Services because it also includes a new data mining component. We also have English Query, Microsoft Repository, SQL Server management tools ­ Enterprise Manager, Query Analyzer, Profiler, and the Agent and Data Transformation Services. If it's data warehousing related, other than the core SQL engine, that's our group's responsibility," says Baker.

"One of the stories I like to tell," says Baker, "involves how we formed our team. It's important to not under-emphasize how much time we spend on building the team, because that's the real leverage ­ if you build a great team, you can build a great product," says Baker. "An enormous amount of my energy is actually involved in team building. Our religion is 'data warehousing for the masses,' if you will. When you think about building fairly complicated software for lots and lots of people, you have to look at the Microsoft Office team. These are the people that ship a spreadsheet that's used by literally tens of millions of people. So one of our key first employees came from the Office team. It's well known that we also seeded the team with a people and technology acquisition from Israel. From that starting six people, we've built this incredibly diverse team of almost 40 people in OLAP and more than twice that in overall data warehousing. You take our goal and economic model together, and it causes you to deliver a different kind of product ­ one that can really be deployed rapidly and with fewer resources," emphasizes Baker.

"It is not simply ease of use that's important," states Baker. "It's also that our solutions are very robust, so our team has been built with people who want to build a really powerful product. But they've got to build it and deliver it in a leveraged way. Our SQL Server/data warehousing products are probably the most open in the sense that we read data from any data source that's out there. Integration in the corporate computing environment is critical. In the beginning, people were not predisposed to believe that our products were capable of integrating into existing corporate computing environments. Obviously, we'd like everyone to use SQL Server, but the reality is that data is everywhere, and with OLE and ODBC we pretty much cover everything."

"Part of my job is to be sure that everyone on the team has what they need and knows what they have to accomplish, rather than how they have to accomplish it. I concern myself more with how we work as a team ­ and making sure that we maintain a sense of humor. Everyone probably thinks my jokes are bad, but we do laugh a lot," says Baker.

"I have always been fascinated with what can be done with technology to help people make better decisions. The ability to support people so they can make better decisions should be available to everybody. I think the Microsoft perspective of volume and leverage and the desire to take on the challenges of finessing the complexity ­ putting it there for people who need it and hiding it from people who would get sidetracked by it ­ are keys to that. This is the best place in the world to do that, and I personally think Microsoft is the greatest software company in the world," says Baker. "It's exciting ­ I still wake up every morning without an alarm clock!"

A graduate of MIT, Baker compares the environment at Microsoft to that at MIT. "When you go to MIT, you are surrounded with people who are smarter than you are ­ faster, quicker, sharper individuals. It is the same at Microsoft. This is the MIT of the software industry. It is amazing how smart people are around here. Every day you get to work with people who are very dedicated, passionate and sharp. There is no place in the world like it," proclaims Baker.


Bill Baker, general manager of Microsoft's data warehouse product unit, leading a team meeting.

Looking ahead, Baker talks about moving data mining into the mainstream. "Data mining is an awesomely useful technology, but there are barriers to mass adoption. What we have done in SQL Server 2000 is address some of the challenges. OLAP is model-driven, and data mining is obviously data- driven. What scares me the most is forcing people to choose between the two techniques. In SQL Server 2000, we've melded the data-driven and model-driven worlds. Now we can do data mining on the cubes or do mining on relational and build dimensions on the cubes. There is a great deal of interplay between the two techniques, and they start to feel seamless. We'll do even more of that in the next release," says Baker.

Baker has a challenge for the software industry. "As I see it, there are three forms of closed-loop analysis. I think we've done a really good job on two, and I want to engage the industry on the third, because it's a big, big initiative. A good model for thinking about this is to look at the IT infrastructure having three big focuses. We have systems to help operate the business, to manage the business and to do analysis and planning. A good example of the first form of closed-loop analysis involves improving an existing data warehouse. We have this capability in SQL Server 2000. If I had a customer dimension that is a million names long, that's hard for a user to navigate. But if I use data mining to cluster that into some logical groupings and build a hierarchy on that dimension, then it is easier to navigate. That's a form of closed-loop analysis. I used an analysis tool, fed the results back into the data warehouse and made things work better. An example of the second type of closed-loop analysis would be where I do some analysis, and find, say, 5,000 interesting customers. I decide I want to send them a particular brochure. In SQL Server 2000, we have Actions, so I can invoke the CRM system or maybe a marketing/promotional system to accomplish that task. That's a form of closed-loop analysis where I've taken results of analysis and fed it back into one of the management systems," explains Baker.

"The first example was closed-loop inside of analysis and planning. The second example was analysis and planning moving back into management," says Baker. "It's going from analysis and planning to operational that's super interesting. Let's say I analyze my customer buying patterns and decide that giving a bigger discount to certain tradespeople who buy over a certain level will really grow my business. What can I do with that information now, other than shout down the hallway? Companies need the ability to perform the analysis, decide on a change in policy, and then with the appropriate work flow and permissions be able to change the discount table in the operational system so the next customer who calls in with a similar order also gets the larger discount."

"The challenge for the industry is how do we take the outcome of analysis and loop it back to the operational side. With all of the e- commerce and e-business initiatives, this kind of capability will become an imperative ­ and it will have to happen without human intervention. This kind of analysis is not just based on a customer ­ it's based on blend of the policies of the company and what is known about the customer. We want to drive this through meta data initiatives like the business rules model that we have, through the business-engineering model that we have and through our partners. We are fortunate to have incredibly high quality partners. We have the partners that work on the front end ­ companies like Knosys, Maximal and Cognos ­ that have done a really nice job delivering data to end users. We also have some really good partnerships with back-end providers who can build the infrastructure that we sit on top of. Our partnerships are very important to our success as a product and a team," says Baker.

"When the analysis results can be blended with company policy, then we'll have super powerful business intelligence. But we can't do it all alone. We want to engage the ERP vendors, the analysis tool vendors and the meta data groups. I'd like to challenge the industry to close that final loop," states Baker.

"I like to say that OLAP was our Act I, data mining is our Act II, and we already know what Act III is, but we're not ready to talk about it yet. But this is our plan ­ to keep providing extremely useful technology that's easy to use. You're seeing us evolve from sort of the cube-centric to star schema to more third- normal forms that are more appropriate for the operational data store and things like that. In the future, you'll see a lot of related things such as message flow in and out of the ODS. You'll also see some really cool things that we're really passionate about on the classic BI side which include adding a real-time component to all of this ­ which is absolutely critical in this e- commerce world," says Baker.

Through focused product development and strategic partnerships, such as the Microsoft Data Warehousing Alliance, Microsoft is committed to ensuring that BI for the masses is not a dream, but a reality.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access