Over the past few years, organizations have realized the importance of effectively tracking and managing customer interactions. Customer Relationship Management systems capture critical information about each individual customer. While CRM has proven to be valuable in building solid relationships with customers, even greater value can be found in the data.
Data mining is a process by which large amounts of customer data is analyzed to find relationships. For marketers, that can mean uncovering correlations between customer characteristics (demographics, psycho-graphics, etc.) and events (purchases, birth of a child, moving to a new home, etc.). Once uncovered, these correlations can inform marketing strategy to drive greater ROI.
Data mining is not a new idea. There are several processes that will allow marketers to dump multiple databases into one master database. Sophisticated computer programs then comb through the data looking for relationships and interactions.
Most programs require a programmer to define specific patterns or relationships to look for, as defined by the marketing team. While this definition provides clarity in the process, it can also miss other relationships that are potentially more powerful.
In our experience, we have found that the data, itself, can provide insights that an analyst might miss. Our process employs genetic algorithms, inspired by evolutionary biology, to help in the feature selection phase. In simple terms, a genetic algorithm automatically mixes, matches and tests new combinations of data to discover relationships more quickly and comprehensively than other processes.
Another advantage of the genetic algorithm approach is timeliness. Since new data is constantly being generated, completing the data mining process quickly assures that it is current and relevant.