October 30, 2007

Database Marketing Consultant

Many of you contact yours truly (Kevin Hillstrom), your humble database marketing consultant, asking for a cheat-sheet that outlines the most popular projects clients ask me to work on. Here is a list of the four projects I am most commonly asked to complete for clients.


Most Popular = Multichannel Forensics Project
  • Explain how customers interact with advertising, products, brands or channels.
  • Illustrate the long-term impact of a catalog marketing program on online and retail sales.
  • Describe how product classifications interact with each other in an e-commerce environment.
  • Outline the ways that online advertising channels interact with each other to grow customer value.
  • Develop a segmentation strategy or mathematical algorithm that forecasts customer segments unlikely to respond to print advertising.
  • A typical project takes thirty days, and often includes a one day on-site to present results and discuss additional analytical requirements.

Second-Most Popular = Catalog Marketing Strategy Evaluation
  • Evaluate contact strategy.
  • Illustrate interaction between catalog marketing and sales in other channels.
  • Develop a contact strategy that maximizes ROI, while growing the customer file.
  • Suggest what might happen if a catalog strategy is discontinued.

Third-Most Popular = Hillstrom's Zip Code Forensics
  • Incorporate your data into the overall algorithm.
  • Provide descriptions of direct marketing responsiveness by zip code.
  • Provide you with a zip code model personalized to your data.

Fourth-Most Popular = Database Marketing Audit
  • A two-day on-site project.
  • A full audit of catalog marketing and e-mail marketing strategies.
  • Grades (A,B,C,D,F ... C = Average) are given for the following combinations:
    • Customer Retention, Customer Acquisition, Data Mining/Analytics and E-Mail/Online Strategy.
    • People, Techniques, Tools.
  • An outline is created for improvement in the areas outlined above, coupled with the expected financial impact of recommended strategies.

Fifth-Most Popular = E-Mail Targeting Algorithm
  • Use past customer purchase behavior, past clickstream behavior, and past user preference information.
  • Develop a scoring algorithm to assign customers to multiple versions of an e-mail campaign.
  • Anticipate a 20% to 25% increase in sales/e-mail for customers eligible for multiple versions of an e-mail campaign.