November 18, 2006

Matchback Analysis

Yesterday's post about allocating online orders to the catalog channel causes one to think about the best ways to measure advertising effectiveness.

Assuming you don't execute "A/B" tests to measure incremental sales across all channels, the matchback analysis provides better answers than doing nothing. However, there are challenges that need to be addressed.

Example #1: You send a catalog on October 15 to your entire catalog housefile. You send an e-mail campaign on October 18 to your opt-in housefile list. You also secure a placement on MSN on October 19. On October 22, a prior catalog purchaser and current e-mail subscriber visits your site with a referring URL from the MSN placement, and purchases merchandise. Which advertising channel is responsible for this order?

Example #2: Same criteria as above (catalog mailed on 10/15, e-mail sent on 10/18, MSN placement on 10/22). Customer visits site after conducting a search on MSN on 10/22, and purchases merchandise. Which advertising channel is responsible for this order?

Example #3: Same criteria as above. Customer visits your website via organic search on Google on 10/24, and purchases merchandise. Which channel is responsible for this order?

What analytical technique do you use to reconcile the issues in examples one through three? If you don't manage this issue in-house, which vendor do you work with to resolve these issues? If you don't work with a vendor or analyze these issues in-house, how do you determine your educated guesses?

Lastly, who are the vendors that you believe are most credible and most talented in this field of study?

1 comment:

  1. Anonymous5:41 AM

    You are asking the right questions. At this point, nobody understands the behavior. Although many direct marketers shy away from primary research, we are going to have to do some to understand what is really going on.

    ReplyDelete