While there aren't too many elements within a paid search listing to test, there are myriad front-end and back-end variables that affect results—from search engine to ranking, landing page to conversion process.
The following are a few metrics search experts find indispensable to determining the effectiveness of paid search programs.
Clickthrough Rate. This measure doesn't provide much insight into return on advertising spend, unless you marry it with post-click behaviors, advises Craig Greenfield, director of local marketing solutions at Chicago-based Performics.
Keyword Source. Marketers should determine the origins of their Web traffic such as search engines, content networks, sources and their distribution options, etc. Some sources might not be worth the effort it takes to manage and optimize campaigns to get the desired result.
Bounce Rate. If this measurement is high, says Greenfield, analysis and testing should be performed to better optimize visitors' on-site experiences for stronger conversions. Two likely culprits are directing traffic to an irrelevant landing page and presenting visitors with an overly complicated conversion process, e.g., a difficult sign-up form.
Conversion Rate. Also look beyond the number of conversions to conversion rates, says Ashley Bruce, search marketing specialist at Spring Park, Minn.'s TopRank Online Marketing. "If the rate is low, you might have a relevance issue between your keywords, ad copy and landing page."
Cost-to-Sales Ratio. "It's always tough to watch averages, because an average can go up or down for a lot of different reasons," says George Michie, principal of search marketing at The Rimm-Kaufman Group in Charlottesville, Va. "[For example,] you brought in a million new visitors to your Web site free of charge, and they didn't convert very well—but it was free. So, yes, the overall site conversion was down, but it was a good day. Ultimately, the metric to watch is the cost-to-sales ratio or, better, the cost-to-margin dollars driven ratio on your paid search program."
Historical Data for Like Periods. As Michie points out, any bid management system looks backward, assessing conversion data for the last few months. "It doesn't know that Christmas is coming, and more importantly, it doesn't know when Christmas ends. Oftentimes, you'll see marketers bidding like drunken sailors into January because the data suggests that everybody's buying like crazy," he explains.
To plan ahead for critical selling seasons, Michie advises marketers to compare sales dollars per click from the prior year or two to current data as it comes in, looking for when the sales surge starts and ends. This allows marketers to do some anticipatory bidding.
The Art & Science of Multichannel Fundraising
Social Media Success