Recently, we excerpted Chapter 2 of Brian Larson's Delivering Business Intelligence with Microsoft SQL Server 2012. Also the author of Microsoft SQL Server Reporting Services and co-author of Visualizing Data with Microsoft Power View, Brian Larson, MCSD and current vice president of Superior Consulting Services, has been in the tech field for more than 27 years, the vast majority of them as a consultant, first in application development and now in business intelligence (BI). Larson took some time to speak to us about what he's learned writing about business intelligence, the unexpected benefits of BI and more.
How did you develop an interest in technology? Was this something you were always into?
Brian Larson: I've always had a tendency [toward technology]. Even in high school, I was one of those kids drawn to the Apple II in the corner of the math room. To be able to create something new in a very short period of time, to enter things in and have a program of sorts materialize out of that, I think that always held a fascination for me. At one point in college, I considered going to seminary, but Greek [language] was my undoing.
Why did you decide to write a book about BI?
Larson: It started through reporting services. I have a coworker, and he had some contacts from previous work he had done with the SQL Server development team. He said, 'Maybe I'll call Microsoft and see if they have any work they want us to do.' A week later, he learned that they did. They were in the early stages of putting together what would be reporting services. He did some coding on that project, and so did I -- I wrote the first version of the PDF-rendering service. That led to some talk with SQL Server Magazine -- we co-wrote an introductory article as that was being released. I thought it was fun. I asked my co-worker if [he would like to] write a book [with me]. He didn't want to, but I sent out some feelers to see if anyone else might be interested. I learned that McGraw-Hill Professional was willing to work with me.
After the publication of the first book went well, the editor asked if I was interested in writing anything else -- since Reporting Services is a part of the SQL Server BI stack, it felt like a natural topic to write about. I also felt there was a need for a book that provided an intro to both BI business analytics and the BI tool set for SQL Server. In addition to this, I was a frustrated science fiction writer. You may notice science fiction references throughout my books -- for example, Galactic Delivery Services, which has next day delivery, same day delivery, faster-than-the-speed-of-light-transport, previous-day delivery!
Tell me something you learned writing this book.
Larson: That my wife is an extremely excellent editor. She also has a tech background, but more from the mainframe side. She came at this project with a technical view, but not an intimate knowledge of the subject. She's very detail-oriented. She has read every word and worked through every example in every edition of each book I've written.
Also, I learned that early adoption of a product is very difficult. One of our goals has been to be the first book on these topics on the market, so we're doing the writing with early, prerelease versions of the software, which makes it harder to figure out how stuff is supposed to work, or we find that [features are] not quite completed yet. We've also had to wait until the release version of the product is available. We need to see what [user interface] changes, etc., have been made so that the final edition of the book reflects the production product.
In Chapter 2, you feature a graph that shows the top of the decision-making pyramid, which is upper management, and the base of the decision-making pyramid, which is forepersons, managers and group leads. How are the priorities of these groups different?
Larson: There are two major differences between them -- timing and scope. [The people] at the bottom -- shift managers, team leaders and so on -- they need data on what's happening right now, how production is flowing, what's happening in the call center. They need to be able to make decisions based on how day-to-day or minute-to-minute operations are going. Their scope is relatively confined to a work group or team.
At the top -- the C-level types of executives -- they're looking over a much longer period of time. How did the organization do last quarter, year to date, year over year? They're making decisions for the entire organization and based on data over time. As a result, their scope [needs to be] organization-wide, and their perception of timing is based more on the long term.
More on business intelligence
Buying a packaged BI system? What to know before you buy
Is your BI system virtualization ready? Find out
Q & A: SQL Server BI meets the cloud
Have you noticed any hesitance around adopting BI in industry? If so, where does this hesitance come from, and how can you convince team members that adopting BI is a good idea?
Larson: Sometimes, you have a subject matter 'expert' who has their set of spreadsheets. They control the data that comes out of their department or area of responsibility. They sometimes fudge numbers a bit to make their area look better. This gets reflected up the chain. There's also control there and a sense of job security. In my experience, this person is often the biggest opponent of BI adoption within an organization.
To combat that, I've worked with those 'experts' and said, 'You acquired this expertise not because you wanted to play with spreadsheets all day, but because you enjoy accounting or production management or whatnot. Let's get you back to that instead of messing with spreadsheets all day.'
What are some unexpected benefits BI can have?
Larson: One of the things that I have seen as we've gone through BI projects is data-cleanliness issues. We've got an order that was ordered for 2102 instead of 2012, state abbreviations that don't exist, things like that. These things go through the transactional system without getting cleaned up. As we're cleaning it up to go into the data warehouse, these things rear their ugly heads. Organizations are realizing they have to clean their data up.
Also, we're finding that different teams deal with data in very different ways: for example, sales and production management systems. Organizations sometimes discover that their processes in different areas of the organization aren't compatible when we bring the data together in a common warehouse repository.
Is there a configuration, version or additional tools from SQL Server that you think is important to have to make the most of BI?
Larson: I'm a big fan of many of the features that Microsoft has added in the most recent release of SQL Server. I would certainly encourage organizations looking at BI products to strongly consider moving to SQL Server 2012. Within that release, the BI and enterprise editions of SQL Server provide the most complete set of BI tools. Also, any organization can benefit from more robust reporting, which can work with any version of SQL Server that supports reporting services. Even other, older versions of SQL Server provide support for robust reporting services.
Where do you think this technology will be in five years? What are some changes to business you expect may happen as a result of better BI integration?
Larson: I certainly think we're going to be much more mobile. We'll see BI delivered to mobile devices -- it's just starting to take off now. In five years, it might be the primary delivery method for BI information. I also think BI will continue to be more and more interactive, moving toward what Microsoft calls 'self-service BI,' where a subject expert will be less and less dependent on a developer from an IT department to pull information out. We'll be able to interact directly with the data.
I think businesses will move toward being more and more data-driven in their decision making. I think there is a lot of 'this is how this has always worked, my gut tells me this is what we should do' today. [Moving forward], I think we'll combine personal intuition and experience with hard data and predictions of what will occur in the near future.