The Human Element Of Business Intelligence

One area of business intelligence that is getting more and more attention is decision making.  We are now capable of processing huge amounts of data with very sophisticated software, but as we have seen in the financial sector (the most sophisticated user of information), there can still be cataclysmic repercussions.  In the November 2009 edition of the Harvard Business Review, Thomas Davenport, the author of Competing On Analytics, discussed some good pointers for those organisations that want to improve how they make decisions.  It is important to start making more good decisions and stop, as soon as possible, making bad, or downright stupid ones.  However, as he points out all too often, decisions seem to come out of a black box.

So the key is to make decision making into a process.  Managers should understand the assumptions which lie behind the decisions that they make, for example house prices will rise by 5%, our chief competitor will lower their prices by 2%, oil prices will rise 20% and so on.  This means that they can be communicated, and if anything changes the organisation can be alerted.  This is vital because you may have built more automated decision making into your processes and you could be left with a torrent of bad decisions as a result.  Sub prime mortgages anyone? Human back up is vital here.

Managers should collaborate.  Organisations such as Air Products and Chemicals have trained their managers to recognise what sort of involvement they should get from colleagues for a particular decision.  Think of the billions of pounds that could have been saved if those who had constructed the financial products that caused the downfall of many of our banks had of explained all the assumptions that were being made.  Similarly, for those who were buying them, and as we found out later, simply did not understand how they worked.

People should actually follow through on the decisions and, as they do at Chevron, have a review of previous decisions and see how they have worked out in the real world.  For this you need a culture that is collaborative and analytic.  Is there beginning to be a sea-change here?

Robert Shiller, the Yale economist, quoted by Davenport, recently said, “You have to be a quantitative person if you’re managing a company. The quantitative details really matter.”

There are software solutions now available that mean that this is all technically possible but as they found at Stanley Works, the toolmaker, you need more than just software.  Here they formed a centre of excellence for product pricing that consisted of  a business unit director, consultants, IT and data mining experts.  The result was an improvement in gross margin that equated to $200 million over six years.

So, in a time of global commoditisation, companies need more than ever to wring out small advantages in assets that their competitors don’t have and can’t easily replicate.  The use of their accumulated data to produce information is one of these.  Another is being able to use this information to make improved decisions that everyone understands and can work on.

Tweet This Post

This entry was posted on Friday, November 13th, 2009 at 10:38 am and is filed under Business Intelligence by Chris Sands. You can follow any responses to this entry through the RSS 2.0 feed.