Q&A: Mesh Old and New Decision-Making Processes to Drive Your Business Forward
As speed of information becomes paramount, the decision-making process must be adapted to meet changing business needs.
- By Linda L. Briggs
- February 17, 2015
Most companies have two styles of decision-making, explains Barry Devlin in this interview. The central BI organization, working on the inside with the data warehouse, has a different perspective from workers on the perimeter of the organization, often using spreadsheets to get their information. "Those two styles exist in pretty much every organization," Devlin says -- it's his belief that the adaptive decision-making process he describes "knits those two processes together in a sensible way."
Devlin, the founder and principal of 9sight Consulting, spoke at a TDWI Webinar in December, From Data Discovery to Adaptive Decision Making, in which he described the business drivers required for speedy and flexible decision making, and introduced the Adaptive Decision Cycle to enable it.
Devlin is among the foremost authorities worldwide on business intelligence. He is a widely respected consultant and lecturer, and is author of the seminal Data Warehouse -- From Architecture to Implementation. His newest book is Business unIntelligence—Insight and Innovation Beyond Analytics and Big Data.
Devlin has 30 years of experience in IT as an architect, consultant, manager, and software evangelist. He is currently developing new architectural models for fully consistent business support -- from informational to operational and collaborative work.
BI This Week: What do you mean by the term "adaptive decision-making process," and what has changed in business to make adaptive decision-making necessary?
Barry Devlin: Let's start with what makes it necessary. We're clearly moving into a new world in which decision-making has become much faster and more interlinked with operations. It's no longer possible for people to spend hours on data discovery before eventually coming up with something useful.
Many of the decisions that we need to make in today's faster, more agile business environment need to be made much more quickly and need to be linked better with processes. We can see that all the way through from operational BI on into operational analytics and beyond.
The basic driver is the speed of business today, but it's a swinging pendulum. When you look at where we were 20 years ago, the problem was consistency of data. Of course, consistency is still an issue, but the pendulum has swung and now the problem is speed. How can I make faster decisions? That's why the decision-making process must be adapted as needs change within the business.
You mentioned earlier that consistency of data is still a problem. Is that keeping companies back in terms of moving toward more adaptive decision-making?
I would say so. Whenever I talk to clients and in classes about consistency versus speed, most people are aware of the issues they still have with consistency of data. They know them very well; they deal with them on a daily basis. What strikes me after 30 years in this business is this: it's not really the same level of problem as we had in the 80s. When you look at what's happened with the more integrated operational systems, I think we've gone to a position where consistency is simply less of a problem.
With the "Internet of things," we're about to go back to a place where data consistency is going to be a much bigger problem. At the moment, we're going through a dip where consistency, at least for much of the standard business world, is much less of a problem than it used to be. For now, business folks are really focusing on speed.
Can you describe what you mean by the "edge-on decision cycle" and then the "center-out decision cycle" that you described in your recent TDWI Webinar on the adaptive decision-making process? Basically, these are styles of decision-making that many companies still practice, correct?
I pondered long and hard about what to call these disparate styles of decision-making. At one level I could have called them spreadsheet-based and the data warehouse-based because that's what they turn out to be.
If you think about it from the point of view of the individual decision-maker, the business problem solver, they are going to be in a cycle that is driven by whatever data they can lay their hands on. You see this with spreadsheets and you see it with data discovery in its more classic form. People grab a piece of data, they play with it, but they can't make the decision they want based on that, so they go around and around. They talk to peers, they find another piece of data, they dig out an old spreadsheet, you name it -- all at the "edge" of the business.
However, a data warehouse, or center-out decision cycle, is really based on the idea that there is someone at the center of the organization who has some level of responsibility for the quality of the data, and some level of knowledge about what it should be like, in order to meet the business needs.
That's why we always have this idea of enterprise data modeling and an enterprise data warehouse at the center of the business. That data warehouse model, the center-out decision cycle model, is one that I brought to the fore back in the 80s, along with people like Bill Inmon and others. That sort of centralized data warehouse environment reflected the need to have data consistency built in. The problem has always been that as decision-makers' needs for data change, the data warehouse struggles to keep up.