By Philip Duplisey – Director Consulting Services
Garbage in, garbage out (GIGO) is a phrase that has been around as long as computers have been around. It has been recently validated for me during the past few weeks when I spent some time onsite at clients for Bardess. What made the experience so defining was that one company understood the concept and was actively working to improve their data quality, while the other company was completely oblivious to it. The contrast made my work with both clients very interesting, and I thought I would share what I learned.
No Business Intelligence tool can substitute good business processes.
Business processes enabled by information systems directly affect the quality of the data in those information systems. If the data quality is poor, it costs the organization money (See Joe’s article on Data Denial), this is a hidden cost, which is not fully recognized until it’s too late (bad decisions made because of bad data). For example if data quality is not part of a sales person’s job responsibility, then we shouldn’t be surprised that we end up with customer master records that are completely hosed. Incomplete addresses, no zip-codes, misspelled customer names wreak havoc on the decision making process when trying to analyze sales performance. At the same time the regional sales manager would have a tough time explaining why sales are down, because “the system” prevented order entry due to incompleteness of the data. Obviously there is a fine balance, but often in the past the balance has been in favor of “ease of use” on the front end with no consideration for what kind of analysis will be required on the data later on.
Dashboards can be made to show you what you want to see.
Mark Twain popularized the saying that there were lies, damn lies and statistics. In the context of business intelligence I might rephrase that by saying there are lies, damn lies, and fancy dashboards! As a consultant using charts and pictures to tell a story, it’s essential to work with a subject matter expert to ensure that the conclusions derived from the charts truly represent the meaning of the data in order to deliver real value to the client. This can be difficult in organizations where the corporate culture is to make decisions based on “experience” and then look for data to validate the conclusion already made. Making decisions based on data is a bigger challenge than what most companies are willing to admit because facing the brutal facts as Jim Collins describes it in “Good to Great” requires leadership and courage.
About the Author
Philip Duplisey, Director of Consulting Services at Bardess, has been implementing enterprise wide solutions globally at Fortune 500 companies for 15 years. He enjoys enabling the organization to find value in their data, and presenting that data in compelling ways that lead to deeper insights and improved performance.
