The Tyranny of OR in a Business Intelligence Environment.
Simplicity is an excellent method of assessing goals, designs and decisions. Every businessman understands how much easier it is for staff when things are straight-forward and clear. Simple is a friend to action and results. Sometimes, however, simple can be confused with easy. Simple is almost never easy. Albert Einstein is credited for defining the appropriate use of simplicity – ‘Everything should be made as simple as possible, but not simpler.’
One example we sometimes see is that we need to choose one intelligence approach OR another. For example, if we are going to create scorecards, some argue there is no need for trending with graphs and other high level information. Many argue that business intelligence is only useful if it is immediately actionable. This argument points out that for busy operational managers and supervisors, that do not have time to ‘analyze’ data – they must focus on getting things done and keeping their tasks on track. Actionable items are those things that effect day to day activity and have concrete actions that can be made with specific results expected. Examples include – missing cash or inventory, hours used above budget, replacing straight time with overtime and using higher skilled staff for unskilled labor. Things that effect profits and can be controlled by managers. Many companies implement tools that will scan data daily and create a list of items to be addressed that allow busy managers to immediately get to work, fixing things that make or save money.
It can be argued that providing too much data can distract a manager from addressing specific items. Hence, if you have specific rules applied then why ‘waste’ time making the supervisor review trending and non-actionable data. This argument states, ‘There are more important things to do.’
A possible ‘AND’ solution would involve an approach that includes a manager with specific responsibilities, clear goals and the ability to manage the allocated resources to best attain results. If such an operation were in place, then the same system that provides the actionable items could easily apply the rules to goals that have been assigned and allow the manager to understand the performance towards those goals. Further, with a collection of this data available, the manager could also receive simple graphs with key data (e.g., sales) over timeframes and graphs comparing this data to relevant past performance (e.g., last week, last month, same period last year). Assuming a manager has clear responsibilities, providing action items along with analysis tools may yield both solutions to the items creating the alerts as well as additional ideas for alerts to keep the business moving towards its goals.
Business intelligence is just that – the combined experience, data and learning of an organization. Important insights today, may become less important overtime as an organization matures. The fun thing about computers is that we do not have to limit our rules and knowledge to just ‘either/or’ choices. We can have our good old rules AND add new rules and even entirely different perspectives. Business intelligence evolves with an organization as it uses it.