Combining CTR & Onsite Metrics for an Adcopy Performance Test

Discussion in 'Pay Per Click Advertising' started by Mscott536, Mar 6, 2014.

  1. #1
    My team is about to start a fairly detailed synchronous PPC ad copy test. We are trying to come up with a metric/formula that would take into account CTR, Pages/vist, and/or bounce rate.

    Another possible metric to tie in would be Conversion rate; however our conversion data isn’t rock solid (see below).

    Here is some insight into our accounts and what we are trying to accomplish.

    We run detailed accounts with large budgets. Our main goal is qualified traffic, due to the majority of sales coming from phone calls( we are currently working on an attribution solution). Also the time from first click to sale can be well over 30 days. These two factors make it hard to run an ad copy test using relevant sales data.

    Our account structure is very granular allowing us to segment different levels of the “buying funnel”. Our hypothesis is that more variables (eg. place, vendor, date) in the KW will qualify the user as being further down the funnel.

    With this in mind, we will be running 5 adcopy variations. 4 of these will be behavior oriented (Research (Learn More) through to Goal Oriented (Buy Now)) and one will be our control which is the best performing copy to date. We expect to see Buy Now convert better with consumers lower down in our funnel.

    We will test all ads running simultaneously for a given period of time(1-2 weeks). Then we will eliminate the worst performing copy and repeat the test. This will continue until finally we are left the best behavior oriented message for each level of the funnel.

    Our issue comes down to; which ad should we eliminate? Using just CTR could be misleading. (eg. high CTR but it may have a higher than normal Bounce rate).

    Our goal in using a blended KPI is to take into account site interaction without having to subjectively decide which copy performed “best”.

    I think the most logical solution would be to weight CTR importance at 60-80%. From there I keep confusing myself with formulas.

    Has anyone used a similar formula combining CTR to onsite data? Or does anyone have any thoughts on good way to go about this? Or maybe even why I’m over thinking this.
     
    Mscott536, Mar 6, 2014 IP