• Database Development & Validation
  We determine if appropriate data is captured and measured.  We inventory our clients' databases and review the data fields for accuracy and necessity.
  
• Primary Research
  We interview selected current customers, former customers and prospective customers to help determine list selection, creative, promotions, packaging, etc. Primary research includes in person interviews, telephone questionnaires, on-line surveys and focus groups.
  
• Customer Demographics Analysis
  We profile the customers and calculate price elasticity, to help increase revenues and determine price sensitivity


• Predictive Analytics
  We analyze historical data to determine trigger events that help our clients market more effectively.  For example, we analyzed public data to determine which customers were more likely to purchase the client's product. When our client contacted the top prospective customers, they were able to immediately close the sales, which immensely improved marketing performance.
• Marketing Resource Analysis
  We analyze statistically all the marketing activities and other factors that may affect sales to determine which marketing is providing the best lift. For example, one client found that direct mail was more effective than mass marketing enabling them to shift their budget into direct.
  
• Customer Lifetime Value Analysis
  We determine the value of customers segmented by a number of variables including profile, geography, etc.
  
• Computer Models
  We build customized computer models to graphically display and evaluate data.
  
• Econometrics
  Econometrics, the use of statistical analysis to determine which mix of factors best explains a result, is both a science and an art. After gathering the client’s data and certain external data, such as economic factors, we build a database for time-series (longitudinal) and/or cross-sectional analysis. The results of the econometric analysis can be loaded into an optimization program to determine the ideal marketing budget and how much should be spent on the various marketing activities as well as the amount on products, on customer segment or by region..
  

We use statistical models, such as linear, log, poisson, integer, non-linear, log log, etc., to analyze the data. Each of these models analyzes the data differently. In addition, each model has lag of one, two, etc. periods. Our goal is to identify the model that best fits predicted outcomes with historical actuals.

In identifying the model with the best fit, we test dozens of various models with lags. The results of each model have several factors that reflect the fit, probability, etc. In addition, we test other factors such as collinearity. The results indicate the lift of various marketing activities and the influence of non-marketing activities.