iSixSigma

Marketing DOE

Six Sigma – iSixSigma Forums Old Forums General Marketing DOE

Viewing 3 posts - 1 through 3 (of 3 total)
  • Author
    Posts
  • #30772

    Lomax
    Participant

    I am working on a Marketing DOE and would like to pose a question to the group. 
    First, let me briefly describe the product and purpose of the DOE.  The product is a mailed offer where customers may choose different levels of an insurance product.  The measured variable for the DOE is response rate (number returned divided by number mailed.)  The purpose of the DOE is to manipulate factors in the mailing such as text font, color, insurance amounts, etc.  For this discussion, let’s keep it simple with a 3 factor, 2 level experiment.  Given that the product is mailed simultaneously, I see no need to replicate the DOE.
    The baseline response rate is 1.5%.  We want to be able to detect a change of 0.1%.  The power would be at least 0.80.  Would you use a 1-proportion sample size calculation in Minitab? I don’t have minitab right now to run the numbers, but let’s say the sample size at 80% power is 20,000.  Do I take the 20,000 and spread it evenly across the 8 runs of my DOE, thus 2,500 mailings at each treatment in the DOE? 
    I would appreciate any insight, feedback, or experience with this type of DOE.

    0
    #80614

    Mike Carnell
    Participant

    Neil,
    Check out a thing called conjoint analysis.
    http://www.surveysite.com/newsite/docs/conjoint-tutor.html
    I think I got that whole thing in correctly.
    Good luck.

    0
    #80650

    Lomax
    Participant

    Mike,
    Thank you for the link.
    Neil

    0
Viewing 3 posts - 1 through 3 (of 3 total)

The forum ‘General’ is closed to new topics and replies.