<< by on October 3rd, 2012
In the world of PPC, we as marketers are swimming in a sea of data at any given time. So, how are we supposed to effectively use this data to make decisions and improve our PPC campaign performance? This SMX East session explores how you can go from drowning in data to a successful Number Muncher (and yes, I am referring to that MS DOS computer game gem) by capitalizing on data analysis techniques to optimize your PPC campaigns.
All of the speakers really harped on the importance of utilizing all available internal and external data and tools to optimize your PPC campaigns. Marc Poirier, Co-Founder and CMO of Acquisio, (@marcpoirier) suggested incorporating the following Google Analytics data to improve your PPC campaign’s customer engagement:
- Bounce rate
- Average page views per visit (or per click)
- Average cost per page view
- Average time on site
- Visits per click
David Rodnitzky, CEO of PPC Associates (@rodnitzky) emphasized the importance of using filters, dimensions, and segments to sift through mounds of numbers to analyze your data on a more granular level.
The presentation by George Michie, Co-Founder of RKG (@rimmkaufman), particularly interested me. Michie explored the issue of paid ads cannibalizing your organic listings as a way to determine whether or not run PPC ads for particular sets of keywords. To determine the potential synergy or cannibalization between paid and unpaid search, Michie suggests calculating the incremental traffic ratio (ITR):
Incremental Traffic Ratio (ITR) = incremental value / apparent value of an ad
The incremental value is the added value that the paid result adds to your organic listing and the apparent value of an ad is the stand alone value of the PPC ad. The resulting number can indicate one of the following situations: destructive cannibalization, constructive cannibalization, or amplification and offer insight as to whether or not you should utilize paid ads for certain keywords.
Destructive cannibalization is just as negative as its name suggests. It occurs when a paid ad deters clicks away from the organic result, thus reducing the value of the organic search listing. When the resulting calculation is negative and suggests destructive cannibalization, avoid running PPC ads for those particular keywords.
In constructive cannibalization, although the paid ad may steal away some clicks from the organic result, it does enhance the value of the unpaid listing. In this situation, the equation would produce a positive number between zero and one hundred and it would not be detrimental to utilize paid search ads for those keywords.
Lastly, in amplification, the paid ad generates its own traffic while increasing the appeal of the organic result, thus increasing its CTR. The organic listing and paid ad exist in a symbiotic relationship and PPC ads should definitely be utilize for those keywords to improve traffic on both listings.
Cannibalization Experiment Methodology
In order to determine whether your paid ads are cannibalizing your unpaid listings, an experiment must be conducted. Just like any other experiment, you must have a control group (those who do not see your ads) and an experimental group (those who will see your ad). This is a difficult feat seeing as search engines are in complete control as to which searchers see your ads or not. In order to bypass this issue, two techniques can be utilized:
- Geographic splits – certain cities are shown the ads and other cities are not shown the ad
- Calendar splits – ads run for a week, day, or hour on and then for a week, day, or hour off
Again, just like any other experiment, you must look out for the existence of external factors that may skew the data. The following are specific external factors to look out for:
- Normal periodic flux in peak days and peak hours
- Seasonal flux
- Promotional effects
- Impact of competitor’s efforts
- Offline marketing efforts