Part 11 in a Series on Web Analytics and Search Engine Marketing Programs
June Dershewitz has scheduled a special Web Analytics Wednesday meeting in Napa right before X Change. If you are coming and are in town on Wednesday, stop by and have a drink. Web Analytics DeMystified, Red Door Interactive and, of course, Semphonic are all co-sponsoring. Drinks and snacks are on us – but June swears she’s closing the tab early enough so that everyone will be okay for Thursday morning!
I’ve actually spent nearly the entire day trying to sort out Huddle assignments for the X Change Conference. It is without doubt one of the most headache inducing and complicated tasks I’ve done this year! So I’m rather happy to be returning to something as simple as SEM analytics…
And this next analysis – measuring conversion by ad creative - is conceptually one of the simplest imaginable. But it is an analysis that's ignored by many SEM practitioners to their considerable cost.
In several earlier posts I’ve talked about the critical importance of using the correct optimization point when doing an analysis. If you don’t, then you are generally worse off after optimization than before. This simple fact is particularly appropriate to the analysis of ad copy.
Unlike keywords, which are now routinely optimized to some measure of conversion or value, many SEM programs still optimize ad copy to traffic. The reasons for this are three-fold. First, Google provides an extremely convenient ad rotation system that lets marketers test creative versions easily. However, it only tests against traffic not conversion or value. Second, unlike keywords, ad copy never show up in your Web Analytics solution without some extra coding. So th analysis of ad copy versions doesn’t just fall out of the measurement system it has to be specially set up. Third, I think many SEM Managers incorrectly assume that the job of ad copy is to simply to deliver traffic – so that conversion isn’t appropriate as an optimization measure.
Let’s tackle each of these in turn.
The fact that any easy system exists for optimizing the wrong thing isn’t an argument in its favor. Google optimizes on traffic for two very good reasons: they have no reasonable alternative and doing so optimizes their revenue not yours. So if you decide you need to optimize on something other than clicks you should just turn off Google Ad Rotations and do ad copy testing manually.
The second issue – the fact that WA solutions report on keyword seamlessly but not ad copy is a practical stumbling block. But it is possible to setup your PPC campaigns so that Ad Creative is trackable.
Which brings us to the crux of the problem – the perception that ad copy is responsible for driving traffic and the web site is responsible for converting it. If this were true about ad copy, it would probably be true about keywords as well. But as it happens, it isn’t true about either. Keywords and ad copy need to be optimized for driving “qualified” traffic. And ad copy is every bit as likely to drive variation in qualified traffic as is keyword.
In studies we’ve done that compared the optimization of ad copy by conversion vs. optimization by traffic (as done by Google rotations), we found that optimizing to traffic wasn’t much better than a coin flip for picking the best ad. The optimization problems are most likely to appear in ads that generate only a modest differential in traffic. Over time, Ad Rotations will strongly favor ad copy that is only a few perctage points better in driving traffic. It doesn’t take much difference in qualification level to more than offset such small traffic differences.
In short, there is no good reason to stick with traffic as the optimization measure for ad copy. What’s more, if you do start measuring ad copy by conversions, you should go back and re-visit old ad copy that you discarded in prior optimizations. Look especially for solid traffic drivers that were marginally outperformed by “hero” versions and ended up being dropped or shown only rarely. There’s a pretty good chance that one or more of these old versions will test better than your current “traffic” hero once you factor in conversions!
Other Posts in this Series: Introduction, Searchnomics Issues, Getting Setup for SEM Analysis, SEM Data vs. Web Analytics Data, High-Level Search Engine Reporting, Analyzing Search Traffic in more Detail, Measuring Search Effectiveness for eCommerce Sites, Measuring Search Effectiveness without Conversions, Measuring Search Engine Marketing as a Channel, Measuring Search as a Channel Part II, Time-Based SEM Analysis, and SEM Optimizations.
Hi,
Very nice informative post on SEO.Thanks for sharing with us..
Posted by: seo consultant | February 06, 2009 at 02:58 AM
Great article, I agree with it. there is no good reason to stick with traffic as the optimization measure for ad copy. What’s more, if you do start measuring ad copy by conversions, you should go back and re-visit old ad copy that you discarded in prior optimizations. Look especially for solid traffic drivers that were marginally outperformed by hero versions and ended up being dropped or shown only rarely. There’s a pretty good chance that one or more of these old versions will test better than your current traffic hero once you factor in conversions. thanks for posting.
-faith-
Posted by: orange county seo | April 01, 2009 at 08:37 PM