Watching the events in Washington and around the world I'm taking the view that my pension is going to be so badly hit that I will probably have to work until I'm 75. The analysis of what has gone wrong will be the subject of reports and books for years to come, but to me the problems may well have as much to do with a failure of information management than with financial incompetence. In a previous blog post I reported on a survey from Capgemini that suggested that 63% of senior managers made decisions at least five times a week without having the right information on which to base their decisions. The current financial crisis tends to support that figure.

For some years I have had to bite my tongue while senior managers tell me that they have all the information they need to make decisions from their business intelligence application. The graphical virtuosity of some BI dashboards make CGI creators such as Pixar seem second rate. The problem with these dashboards is that they tend to become a self-fufilling prophecy, showing established relationships between business variables. But we have not lived in established times since at least 9/11. Changing the relationships is not just a drag and drop exercise using a wizard but requires analytical thinking about how valuable and predictive these relationships are going to be. For years one financial institution in the UK used to get its management accounts seven weeks after the month end. Earlier this year the business model for the institution collapsed during that period and the Board had no visibility of the disaster that was overtaking them.

The second issue is the volume of data and of documents/emails etc that are being generated inside the enterprise. Most companies have no idea of the volume of information on shared drives, but my guess is that for most large companies it is in multiples of terabytes. The folder structures of these shared drives is usually noticeable by its absence. Whether it is managers looking for information or compliance authorities checking on the state of the business searching through this black hole is all but impossible. Even powerful enterprise search applications are not going to do much to help. We are in to text mining and sentiment analysis where context is everything.

But even then this is not just a technology issue. Companies still pay little attention to information management. No one wants to take responsibility for information quality, conformance and discovery. You can see this in the intranet space where often it is the department that has the least effective way of saying 'no' that ends up with the intranet by default. Ask any Chief Information Officer what percentage of the information assets by value to the business they have under their direct management and be surprised if they have any sort of answer. Few companies seem to take an integrated approach to information management and according to a Deloitte study from 2007 (pdf) there are some serious deficiences in the way in which non-financial information on business performance is provided to Board executives.  In October the Butler Group are presenting a two day course on information management and I would absolutely support the introduction to the course. Realistically I doubt that having an information management strategy would have prevented the credit-crunch - there are none so blind as will not see. But the scale might have been reduced and the recovery enhanced by a more professional approach to the management of information.

Martin White