Just like economists and financial analysts, search engine marketers like models, the simpler the better. And just like economists, marketers are often times led astray by these very models. From the very start the focus was on getting the searcher to click on your listing, visit your website and convert. Any tracking began with the keyword used and, hopefully, ended with a conversion (sale, download, newsletter sign-up, etc.). This ability to measure results gave advertisers unprecedented insight into their customers’ behavior.

Unfortunately, search marketers, and the advertisers they served, were living in a bubble. Although a very special advertising medium, search is certainly not the only one. Some keywords used are so specific that a searcher had to be exposed to a brand, a product or a service before getting to a search engine. Marketers also began to notice that the search volume would at times increase after a display or a TV ad campaign was launched. Clearly, different advertising channels were influencing each other, in what is known as the cross-channel effect, but in what way and to what extent?

The First Steps In Source Attribution

One of the first to notice this gap in measurement were Google’s rivals, Yahoo! and Microsoft. Since Google enjoys close to total dominance in the search space but lags behind when it comes to display advertising, both players began to push the idea of source attribution. Not that long ago Yahoo! Search Marketing introduced their assist metric which attempts to figure out which other channels might have assisted the keyword in question.

Microsoft went even further by introducing engagement mapping to their Atlas tools, which, just like the assist metric from Yahoo!, attempts to distribute credit beyond the last click. Here is an interesting research paper on engagement mapping (from Microsoft). It is likely that both Yahoo! and Microsoft are trying to take some credit away from the last clicks, which are overwhelmingly delivered by Google, and attribute it to other channels, like display, where they have a much bigger presence. However, this doesn’t mean that the cross-channel effect is not real.

Where Are We Today?

I was fortunate enough to attend the Search Engine Strategies Conference in San Jose last month, where I sneaked in into one of the Advanced PPC sessions, called The New Search ROI: Measuring More than Conversion (click on the link for a great session coverage by the Bruce Clay team). It was an eye opening experience as speaker after speaker highlighted the gaping holes in the current understanding of source attribution.

The technology to track multi-channel effects and to measure multiple events that led to a conversion is emerging (Atlas’ Engagement Mapping and Mediaplex’s Path to Conversion are two examples). Using cookies you can track the same visitor as she clicks on your banner, browses your site, leaves and then lands on it again through Google a few days later. What is even more difficult is to organize all of these events in order of importance. Is the last click the most important? How much credit should be given to that first display ad interaction?

Our understanding of customer behavior is constantly evolving. While the Internet offers an unprecedented level of control over the results of your marketing campaigns, there are still gray areas that remain to be explored and better understood. It helps that large players like Microsoft have a vested interested in helping advertisers better understand their conversions and attribute credit to the right sources.

No related posts.