Joel Cadwell
Like many in marketing research, I spent my first years using SPSS and SAS along with proprietary programs from academia and private firms. Then, I discovered R. I tried it because it was free, but I switched to R because this was the platform used by so many innovative researchers across a wide variety of fields. R is more than just statistical software. Along with the R code, you get thoughtful solutions to complex substantive and statistical problems.

For the last six years I have been using R packages for all my marketing science consulting. For example,
  • hierarchical Bayes' choice models (bayesm) for product and service design
  • ensemble clustering (clue) and model-based clustering (mclust) for segmentation and target marketing
  • archetypal analysis (archetypes) for profiling customer types given high-dimensional multivariate data
  • Bayesian networks (pcalg, bnlearn) and graphical models (qgraph, gRain) for uncovering causal connections and displaying relationship networks
  • predictive modeling algorithms from statistical learning theory (randomForest, party)
  • psychometric scaling including latent trait modeling (ltm) and multidimensional item response theory (mirt)
  • biplots (FactoMineR), correspondence analysis (ca, anacor, homals), and multidimensional scaling (smacof) for graphic displays of market structure
  • partial least squares path modeling (plspm, semPLS) and structural equation modeling (sem, lavaan) for causal modeling
  • plus all the basic R statistical functions for multivariate statistics.
After receiving my PhD in statistics and measurement from UCLA, I taught at Rutgers University and published original research in behavioral decision theory. But I was drawn to the private sector by the opportunity to work in the marketing science group at BBDO/New York. Returning to Los Angeles, I continued working in market research departments for a large regional bank and a major PC software publisher. Finally, after a couple of years with a full-service marketing research supplier, I started my own consulting firm to concentrate on marketing science modeling.

Except for teaching a few MBA classes in marketing research at UCLA, I have spent the last 20 years as a consultant working on a broad range of strategic marketing issues. My clients have included automotive (Honda, Nissan, Toyota), software (Microsoft, Intuit), hardware (Cisco, Intel, IBM, Xerox, Hewlett-Packard, Lexmark, Motorola), banking and financial (Citibank, Wells Fargo, B of A, Fidelity Investment), health care (Blue Shield of California, Kaiser-Permanente, WellPoint), telecommunication (AT&T, Sprint, Pacific/Southwest Bell, GTE/Verizon, Qwest), retail (Albertson/Vons Supermarkets, Campbell Soup, Dryers), restaurant (Jack-in-the-Box, Pizza Hut), and satellite/cable providers (DirectTV, Comcast) plus the Los Angeles Times, LifeLock, Tumi, eBay, and Alaska Airlines.

No comments:

Post a Comment