The Science of Sample Sizes


The Science of Sample Sizes

Perry Drake of Drake Direct helped co-write an artilce for Target Tipline to assist direct marketers in designing and analyzing their marketing tests. This article appeared in the July 20, 2005 online publication

The secrets of testing come down to seven straightforward and essential guidelines:

  • Ensure the sample sizes are random and respresentitive
  • Block sample sizes when appropriate
  • Include a control panel in the test plan
  • Back test package changes to the control
  • Test only one change at a time to your control
  • Test only for meaningful interactions
  • Ensure adequate sample sizes

Of all these rules, it is the last one–adequate sample sizes–that poses the greatest challenge for many direct marketers.

 Click here to download the complete article in pdf file format