Hi Tom,

Great post. I couldn't help but think of the contrast between standard business decision making (quick but fairly uninformed) to standard experimental process (slow but more likely to form justified conclusions).

Perhaps businesses should think about introducing aspects of experimental design into their decision-making process, and specifically the need to quote past decisions or empirical data as justification?

I've posted a more detailed discussion as well.

I have noticed how many managers asking for decision support dashboards to their transactional systems, but fail to identify key information they would like to monitor on these screens or reports. If these analytics tools were implemented in a meaningful fashion, would it not be a step in a direction of decision making systems? It seem the managers are overwhelmed by technology tools available. Most cannot articulate what they need to know to make pro-active decisions.

Tom,

Boy are we ever on the same page. Today's BI solutions are largely about informing people, but the synaptic void between information and decision-making is not addressed. There is a pyramid definition that is ubiquitous - separating strategic, tactical and operational decisions and aligning different BI technologies with them, but in fact, this model is not justifiable.

I don't agree with Mr. Bounds. Experimental science is all about conclusions, not decisions, as the latter are never presumed to be final or even verifiable. There is a huge amount of work to be done here, but we can stand on the shoulders of giants like Herbert Simon who looked at decision making decades ago.

So BI may be useful for prioritizing what decisions to make, and informing those decisions that are driven by data, and evaluating the outcomes, but the understanding of decisions themselves is still an unplowed field.

-NR

Dear Prof. Davenport,

For Information Systems to have relevance in an organization, it may be critical to complete the input-process-outcome-feedback-control loop from a system perspective. More often than not, solution providers emphasize on the input and process part, and occasionally on the outcomes part. Very rarely do organizations or the vendors of solutions worry about feedback and control.

Yet, this cycle of forget (the undesirable), borrow (the desirable) and learn (a continuous process), to use the analogy of Prof. Govindarajan and Chris Trimble, is essential if we are passionate about improving the quality of decisions. Although management may never reach the level of a pure science in terms of being able to stand the test of repeatability, a constant endeavor to look at processes and their outcomes may still prove to be a valuable tool in ensuring, at the very least, that we do not make the same mistakes again and again. And that, in the words of Akio Morita, would itself be a great accomplishment.

It is time for organizations to look at IT not as a fad, not as something to be implemented because competitors are doing it, but to treat IT as a valuable facilitator to improve the way we work, the way we decide, and hence the quality of life itself.

Warm Regards

As usual, Tom Davenport has identified a key issue and provided some sharp insights. Questions for taking the discussion further might include:

What do we know about information availability, use and decision making? In fact, there is a fairly rich run of literature, but a good deal of it is in information studies journals rather than management journals and therefore disconnected from management discussions. Of course, we need more studies. But we also need a framework, e.g., what are the criteria for a system that is really useful for decision making?

What do we want to know from decision-makers? Just asking them what information they had and used (or didn’t have and needed) would be only a first step. We need a series of probing questions that draw them out and, for instance, get at the issue of intuition/gut feelings vs. decisions based on objective information.

Who should be in charge? The responsibility for taking the lead in designing information systems to meet decision making needs should usually fall to institutional Chief Information Officers. But, too often, they are bogged down with information technology problems rather than focusing on strategic information management issues.

Tom,

"We have lost much of the connection between the supply of information and the demand for it in decision-making." I agree with the statement, as often, the people who are making decisions are not the ones who help to design the information systems that are put in place. Even worse, operational management will leave the decisions to the IT department, and have little input in the development process. So what happens, is you have IT create a information system, that is probably too complex, and will be developed from stage to stage with little insight from management. With the management on the outside looking in, perhaps only having the IT team give them reports on progress and costs, the development goes forward unchecked. Then the project is completed, any new systems go live, and front line managers are given their, "new powerful tool" to use. But the problem is that the system is overly complex and probably can't do the simple stuff that, if the front line managers were asked, would have said they needed. That's why innovation teams need to be created from a diverse group of individuals and not just C-level employees, and it's why IT should not be the ones making IT decisions.

As the former head of Information Systems of a $ 3.5 BN organization, I cannot agree more with Tom's basic posit about the disconnect between the supply of information and decision making. I also agree with the essence of what Neil is saying. Information systems try to automate 'what we do' and not 'how we think'. The decision-making process is all about 'how we think' and cannot be codified beyond a point. That's the primary reason why all BI systems have failed to deliver on their promise of enabling better decision making. At best, BI can be a powerful medium of making information visible from different angles and dimensions...but that doesn't aid the making of good decisions, especially when a lot if it has to come from the intuitive right brain. The promise of automated decision making may be possible if artificial intelligence based systems become a reality...where you teach the application to mimic the way departments, roles or organizations think, but isn't that far off ?

When you say we might need to "start an inventory of key decisions" you are on to something here. We at Austhink Software have been keeping inventories of our decisions in an online repository. The key to doing this has been to use a decision visualisation tool bCisive http://bcisive.austhink.com (which we happen to make - hey we "eat our own dogfood").

We don't document/inventory all our decisions of course, but important ones esp. those needing collaborative input are - and we have found it very useful.

Tom,

I enjoyed your article about decision-making. Many times management is scared about what decisions to make for fear of the consequences they face, especially in the current market with more competition and globalization then ever before. Organizations and management should ask themselves what the organization needs to grow and develop.

I recently read this article on this website that helps to lay a processes for managment to brainstorm through to understand that and how to implement it for lasting results. The employees are so important in making those decisions. You should check it out:

http://www.organizedchange.com/decide.htm

Jessica