Q&A: Eckerson Book Delivers Insights from Analytics Leaders
Wayne Eckerson's new book, "Secrets of Analytical Leaders," weaves insights from the well-known BI expert with comments from top names in the industry.
- By Linda L. Briggs
- January 22, 2013
Wayne Eckerson describes his new book, Secrets of Analytical Leaders, as "a panel discussion in print, with me moderating the seven smartest people I know in BI." In this interview, he discusses the insights he gathered from top experts in the industry on delivering BI and analytical solutions.
TDWI: What led you to write a book on the Secrets of Analytical Leaders?
Eckerson: It's interesting how the book came about. I was talking with one of the people whose comments are included in the book, Eric Colson, formerly of Netflix -- someone who's probably one of the more original thinkers in our industry and whom I've always enjoyed speaking with -- and he said, "You know, Wayne, why don't we collaborate on a book? I have all these great ideas, but I know I'll probably never get an opportunity to write them down." I said, "That sounds like a great idea." I put it on hold because that's a big project, right?
Then I started to think that I've met a lot of people like Eric in the industry over the years, people whom I've had speak at [TDWI's] BI Executive Summits, and people who've won awards, and people who I've interviewed. They all have great stories to tell, and they all have great ideas -- there are lots of things that I've learned from them. I thought, why don't I write all their books as well? That became the seed for the book.
So you drew on your many years and contacts in the industry?
I took some of the people that I've known over the years from different industries and did three-hour to five-hour interviews with each of them. I basically asked them one question -- "What is the secret to your success in writing a BI analytics program?"
At that point, I realized that I actually had too many people -- too much information, really -- so I narrowed it down to seven people. I picked people who had responsibility across data warehousing, BI, and analytics -- a sort of triumvirate of capabilities in their organizations, more or less. I interviewed them, asked them the question, transcribed what they wrote, and then fed that back into the book word for word.
You make it sound simple, but there must have been a lot of judicious editing. Also, one thing that adds to the book, I thought, is your comments in between their commentaries.
Basically, I let them tell their stories in the book and I wrapped my perspective around it. One reason it took so long is because originally each of them had one chapter. As it developed, I realized that that wasn't so interesting, so I decided to break it up into topics, where each topic would be a question that I get asked about analytics.
It turned out to be sort of a panel discussion in print -- me moderating the seven smartest people I know in BI, and getting their opinions on key topics about delivering BI and analytical solutions.
I had fun. I hope the audience gets as much value out of it as I did.
What are some key ideas that emerged from the book?
If I were to step back, right off the top I'd have to say that one of the themes that came out of the book would be speed-to-value. That's a key one. You have to deliver value fast if you want to succeed.
Another theme is, you have to get the best people and you have to embed them into the business. Create one team, not two teams -- you don't want a business team and a technical team. You need a single team, so they're all talking the same language. That's kind of an über-theme that comes out across the entire book.
Another main theme is, you have to get your data in order to do analytics -- it's the old garbage-in-garbage-out story.
One thing that surprised me was a comment you made early in the book, that as you spoke to these people, each of them hesitated when you asked them to define the term "analytics," despite being at the top of this field.
It's often hard to define a term in our industry, just as companies have a hard time defining exactly what a customer is or what a sale is.
When everybody uses a term, which is what happens with analytics, it tends not to be clearly defined, considering all the perspectives and agendas attached to the work.
In the end, though, in the book, everyone all basically had the same definition, which is a kind of big-picture definition of analytics.
Then there's the technological definition of analytics, which has more to do with analytical tools for reporting and now for mining. I think one reason they hesitated is because analytics is an umbrella term that is both a business as well as a technical term. I asked them for only one definition, not two.
What's your definition of analytics?
I described it in the book as "everything involved in turning data into insights into action."
As I say in the book, it can be a big "A" versus little "a" thing. Analytics with a big A is the umbrella term that represents everything -- the people-processes-technology required to turn data into insights into action. Analytics with the little "a" is all the technology and tools needed to do reporting and analysis. Advanced analytics typically refers to the data-mining tools -- machine-learning tools, more significantly.
I found it interesting that Eric Colson said that although analytics is about measuring, it's surprisingly hard to find explicit metrics about the contribution of analytics to a company.
It's quite ironic, isn't it, that we're the ones who measure everything but we don't have a measure for our own success?
Eric was arguing that the best measure of success is how often you get mentioned as a group in the executives' briefings, or monthly operational meetings among a line of business heads, or the executive committee if you're relevant enough that people would say, "Yes, this group enabled us to do (this) ... and they enabled us to do (that)." Of course, that's still a very subjective measure.
Another interesting comment in the book was from Tim Leonard, who has worked in military intelligence. He said he learned the hard way that you can't be perceived as an IT person out there. You have to be seen as a business person who understands how to use IT.
Yes, that goes to the fact that you have to be one team. You can't be an IT group at its own agenda. You have to be viewed as a businessperson with technical knowledge. That way, you gain credibility and you gain context, so you know better what they need and how resources and capabilities you have can further their goals and objectives. Yes, as soon as you start talking techie, you start divorcing yourself from the business and creating a gulf between the groups, and that's not healthy. It leads to a division that sometimes can be hard to bride.
Was it hard to write a book about something that's changing so very quickly?
Yes. Last year, analytics was really hot. Now it's big data -- analytics had its two years in the sun, then it got replaced. The thing is, big data is useless unless you analyze it, so it really goes hand-in-hand – big data and analytics. I probably would have been better served to call the book Secrets of Big Data Analytical Leaders.
Well, maybe the next book ...
Big data will go away, too. It's just data and it's just about reporting. All we do, really, is monitor and analyze data. Some people can make smarter decisions and better plans, and optimize performance. In a nutshell, that's what we do. So in a way, it wasn't hard to write about these things at all because much of it revolves around core principles that don't really change.
One thing that was interesting to me was talking to Kurt Thurling who is the token statistician among the group, although Eric Colson was a trained statistician, too. Kurt didn't want to talk about algorithms or developing models. He wanted to talk about curating the data -- getting the data ready to be modeled by statisticians. He was also interested in how, once you develop the models, you then put them into production. He talked about how you do that in a big company with lots of models and lots of regulations about how you deploy those models. I thought that was interesting.
I also thought it was interesting that despite this move to big data, and even though quite a few of these people are using Hadoop -- despite that, there's a big, big, big emphasis on data warehousing. That's why part of the book emphasizes that the data warehouse, even in the age of big data, still plays a critical role in helping companies monitor and analyze their data.
It was interesting hear that data warehousing is still so important to these big companies.
Yes, but you have to believe that the data warehouse will definitely be sharing the limelight with Hadoop going forward. If the infrastructure of Hadoop evolves to become more real time, it could become a central hub for data management.
What else surprised you as you conducted the interviews? You mentioned Kurt Thurling's focus -- were there other things?
Kurt – yes, that surprised me. I thought he'd begin to talk about developing models, but he wanted to talk about the other things.
He talked about curating the business.
Yes, and about just how essential the data warehouse is to certain organizations -- Internet companies that you wouldn't expect that of. Zynga [Ken Rudin's former company] bypassed Hadoop entirely to move data into a data warehouse because it knew ultimately that the questions they wanted to ask were dimensional in nature, and the best way to ask dimensional questions was through SQL-based tools. I found that interesting.
Another interesting thing that came out in my interviews was the fact that these people really didn't want to talk about tools and technology as much as people and processes. The key to their success, really, lies in getting the best people -- then in organizing them the right way and motivating them, so that they became highly productive and enjoyed what they did. That's why so much of the book is about people, process, and projects. That's where the people I interviewed really wanted to focus.