Bill Baker spent more than 12 years at Microsoft where he started the Business Intelligence (BI) team within the SQL Server product unit. Baker and his team delivered three releases of Analysis Services, Reporting Services, Integration Services and the SQL Server Management Studio. During that period, Bill was a frequent spokesperson for Business Intelligence at Microsoft.

His expertise and work in leading teams to design and develop Microsoft solutions resulted in his recognition as one of only 36 worldwide Microsoft Distinguished Engineers, an accolade bestowed by the company to technical contributors whose high level of performance, technical vision, expertise and world-class leadership are instrumental in developing and driving products and standards for Microsoft and the technology industry into the future.

Bill’s career includes more than 29 years of experience in enterprise computing and data warehousing product development at Oracle/IRI Software, Softbridge Microsystems and Interactive Data Corporation.

He holds a Bachelor of Science degree in electrical engineering from the Massachusetts Institute of Technology and is an industry-recognized authority on Business Intelligence who frequently spoke and wrote about BI on behalf of Microsoft.

Bill was CTO of Visible Technologies, which produces software to measure and analyse brand perceptions and performance in social media. Baker currently serves on several corporate and advisory boards and assists several local software start-ups in the Seattle area.

---

RM:

Microsoft introduced Object Oriented orthodoxy to DBAs. It may have seemed obvious to people at Redmond but alien to the majority of DBAs. Do you think was a smart move?

BB:

One of the things the team realized a while ago is that developers are a key constituency. SQL Server 2005 really brought the database and development worlds together in a new way. Developers love all that orthodoxy. But it does not define a single way to manage and administer the database. The DBA community can work in the traditional ways. Even there, you see people in that group adopting, and liking, some of the new methods as well.

RM:

What were your favourite features SQL Server, and why?

BB:

Well SQL Server 2008 was really rich; there was a lot to like. From my perspective working with unstructured data, I spent time learning Filestreams. The ability to have the goodness of a database and have huge text volumes is really interesting.

RM:

You directly influenced SQL Server in so many ways. Can you tell me a couple of examples about features you have lobbied for, and were eventually added specifically because you requested and lobbied for them?

BB:

Well, you could start with Business Intelligence (BI). When we started the OLAP project, the idea that a relational database would have Business Intelligence capabilities was really new. We grew from a team of one person into a very solid, well rounded team with rich set of features. Over the years Microsoft has become a true leader in Business Intelligence.

RM:

Business Intelligence has been around in one shape or another for at least a decade. Why do you think it has taken corporate users so long to realize its potential? Do you think BI tools have been too esoteric?

BB:

When I started in Business Intelligence, it was truly an industry of specialists building specialty software for other specialists. So yes, for a long time, the tools catered to the experts and were too esoteric for most people.

The other side of that equation is that while the specialists go to work in the morning to ‘do BI’; most people go to work to do their jobs. Most jobs in most companies require no BI or very, very little. So you combine a broad population with minimal needs with heavy-duty, rich tools, and you get a slow adoption curve. But that’s changing, rapidly. Look at what the Excel team did in Office 2007. They are doing a lot to mainstream business intelligence in a tool so many people know and love. That does a lot to accelerate the adoption of Business Intelligence.

RM:

A number of BI decision makers are still grappling with the question of ‘Build or Buy’ – should they build the BI application from scratch or buy a pre-pack application? What would your advice to them be?

BB:

I think the build-buy pendulum has swung pretty hard toward the buy side in just about every segment. Small companies don’t generally have the means to build. Medium companies are generally fine with pre-packaged solutions; meaning they will change their business process to accommodate the software solutions they buy. It’s in the enterprise that you see a hybrid build-buy.

That is, companies will buy packages and then spend ten or more times the license and maintenance fees to customize the software to support their business process. I wonder if they aren’t getting the worst of both worlds. Over the last year I’ve heard many CIOs say, essentially, that’s it, we’re buying a packaging and using it as it comes. I think there is an opportunity, if not a call, for the LOB vendors to shift the effort from expensive customisation to more practical and useful configuration.

RM:

In the last few years, the biggest software deals have the big enterprise companies were buying up business intelligence companies: SAP (Business Objects, $6.8 billion), IBM (Cognos, $4.9 billion) and Oracle (Hyperion, $3.3 billion) but according to the New York Times, tech spend in the US is down. How badly do you think the downturn affected Microsoft’s BI division and that of other companies?

BB:

BI is still a CIO and CEO priority. So while the overall downturn might impact growth, I think you will still see positive growth in the segment. The last big storm for the BI market was the concurrent impacts of the dot-com mania and the Y2K panic. Both took spending away from other initiatives. Even then, BI weathered the storm. It will again.

RM:

Search engines are revolutionizing the way people do business and access information. More than 60 billion searches are conducted online each month and 90 percent of online consumers use search engines as their primary source of information. With Facebook, Twitter, MySpace and other social networking sites encouraging us to share our personal information, is social engineering going to become easier to do? What will security look like in a world where nothing is private?

BB:

First, search as a BI tool is really exciting. In some ways, it’s end-user Extract Transform Load (ETL). If vendors make it easy to search for information and drop that info into a tool, users will rejoice.

In general, social media is about transparency, not about loss of privacy per se. It’s up to the sites to protect privacy while allowing the members to share more transparently among the community. I think in general, the generations that are most transparent in their lives and on the internet, are also savvy about the perils of the internet. In the short-term, while other generations, such as my generation and my parent’s generation, get used to social media, I think there is reason to worry. I have seen some good attempts at education and at safeguarding privacy on many of the social networks.

RM:

Search Engine Reputation Management (or SERM) tactics are often employed by companies and increasingly by individuals to shield negative stories about them. Is it fair to sanitize this sort of content? Do you think it helps the ethical reputation of the web?

BB:

I’m not an expert at SERM. My understanding is that it is more about balance than it is about sanitizing. What’s on the web pretty much stays there forever. SERM doesn’t remove anything; I believe the goal is to ensure that older, less favourable news gets balanced by more recent good news, assuming there is some. I think product reviews are a good example. While the recent UAL situation didn’t arise via search engines, it does show the need for companies to actively monitor news about themselves and their reputation.

RM:

Microsoft has introduced social media monitoring services from multiple suppliers such as TruCast from Visible Technologies to build a community around products. Commuinity led software development rather than company software development is becoming the norm. Is it fair that customers would be expected to pay for software that has been largely developed from free ideas?

BB:

I think there is a difference between community led and community influenced. I think every company benefits from a more intimate relationship with customers and potential customers. In some ways, that’s the origin of Business Intelligence. It was about customer intimacy. But it was always hard because there was no good proxy for customer attitudes and intent. Now, using tools like TruCast, companies can directly listen to and engage with their customers. So, yes, I think companies should listen, learn and evolve based on the community’s input.

And as a customer, I’m happy to pay for products that are better because of that input. To the second part of the question, and specifically to software, if the community truly develops the product, then I think the economics change. I still might pay for the aggregation and testing of all that input, and the support. But I’ll happily pay for software that a vendor wrote and improved because of community input.