Modeling

Like a hybrid of science fiction and fantasy, modeling uses advanced mathematical equations, along with other data, to predict the future. The wielders of the crystal ball are our highly detailed, laser-sharp analytics team. Through precise modeling techniques, they can gain the foresight to react to our clients’ needs before it’s too late. So they can answer questions like: Should you spend more? Spend less? Maybe target a smaller group?  Will someone pick a product? Which one? Is it time to start over?

The techniques we use are cross-channel — DM, TV, online, etc. — and include: linear regression, logistic regression, multinomial logistic, and CHAID.

Analysis

Over time, results from our modeling have provided us with a wealth of information, ensuring we’re constantly better equipped to respond the next time we’re faced with a similar challenge.

Through our learnings, we’re able to analyze a number of different items, such as:

  • Seminar placement
  • Effects of repeated targeting
  • Data source evaluation
  • Postmortem campaign reviews
  • Model validation
  • Creative test design

Tools

Our in-house programmers use three of the industry’s most known programming languages: SQL, R, and SAS. This benefits both us and our clients because our team isn’t limited to what would be offered in a standalone tool — using these three languages allows us to work directly with the data. Additionally, because these languages are so widely used, we’re able to share code directly with any in-house analytics team a prospective client may have.

Our objective is simple — in order to give our clients the most accurate forecasting and modeling possible, we utilize the best technology available.