In depth customer knowledge drives new business – acquisition, cross sell and up sell. However, failure to correctly apply firmographics, corporate hierarchy or purchase history can leave your vision fuzzy. Massini Group customer database consolidation and predictive analytics will add clarity.
The aphorism "know thyself" has been attributed to numerous ancient Greek sages, with the most well known being Socrates. Said to refer to the ideal of understanding human behavior, here we refer to it in the sense that understanding what has made you successful with your current customers will make you more successful with your future customers.
There is definitely a large amount of specialized knowledge, craft and skill associated with doing so successfully. A great deal of good can be done with any of the growingly common forms of statistical extrapolation or predictive analytics such as :
- projected account spend / profitability ;
- propensity to buy / respond / be the decision-maker ;
- touch combination correlation to sales success ;
However, Massini Group has rescued numerous strategic targeting, segmentation and campaign management predictive analytics projects that did not pay enough attention to the accuracy of the process by which firmographic data was appended to a customer file to be used as independent variables. False positive matches, duplicate matches at different levels of the corporate hierarchy yielding wildly different attribute values or simply inadequate levels of data population can yield results that are downright nonsensical.
The key to successful predictive analytics is unflinching attention to detail in the initial phases of the effort. Massini Group insists on deterministic validation of all appended data from commercially available business reference sources and a minimum input data population rate of 90%. Massini Group commonly achieves these levels of accuracy with a combination of commercial vendor matching with after-the-fact match validation and manual research and assignment of previously unmatched records.
The second most common pitfall in the use of predictive analytics is inferring more detail than is truly available. Massini Group asserts that simpler is always better. Highly complex multivariate models are highly sensitive to inherent bias in the input data set. Further, when they are applied to a comprehensive marketing universe for the purposes of providing strategic direction, limitations to the data population of that larger data set can result in fundamental breakdown of the process.
Another class of highly effective predictive analytics is based on simple correlation analysis based on solid and auditable codification of variables with large numbers of values, such as abstracting contact titles to organization function and level or coding campaign responses into simple message directions and relationship stages.