Test and measure search campaigns

Published on .

Kim Ann King is CMO of SiteSpect Inc. (, a multivariate testing and optimization company. BtoB recently asked King about best practices in testing Web site content. BtoB: How can testing landing pages benefit search marketing campaigns? King: Online marketers rely on search engines such as Google, Yahoo and Bing to bring visitors to their sites, but that is only the first step. Once visitors arrive at a site or landing page, companies have to keep the momentum going and, ultimately, convert those prospects into buyers. Testing and targeting enables marketers to customize and optimize online content, such as offers, landing pages, graphics and copy, depending on what seems to work best. The process also allows them to perfect their “context,” such as navigation and conversion paths. The results of these efforts help marketers understand what is most appealing to visitors who come from their search campaigns. BtoB: Can you give some examples? King: Targeted content allows visitors to quickly and easily find what they are looking for on your site. For pay-per-click campaigns, landing page URLs can be specified for each ad, ad group, keyword or creative. Testing is simply the evaluation of the impact of combinations of factors and variations, revealing significant interaction effects that can have a dramatic impact on a Web site's conversion goals. For example, we've seen companies increase conversions by as much as 71% by first testing page layout and content and, in response, streamlining landing pages, simplifying form fields and targeting content variations based on keywords. To do this, companies can use a “test-learn-repeat” pattern, typically testing such elements as copy, font/color/size, buttons, navigation, images, layout and functionality. BtoB: What are some measurement best practices? King: A Web site testing program should integrate and collaborate closely with your overall Web analytics and search marketing efforts. Metrics and key performance indicators as part of a Web analytics program will probably dictate, to a large extent, which tests should be undertaken. Next, you want to clearly define success and failure. Success can range from obvious financial gains, to increased user engagement or fewer support calls. But even a “failed” test is helpful, because you'll better understand what didn't work. The only real failure is a test that is not properly designed or carefully executed, or one that is done within a vacuum and without executive buy-in. —C.H.
Most Popular
In this article: