• by Rob Garner , May 27, 2009

At the Search Insider Summit in Captiva, Fla., Gord Hotchkiss hosted a panel and clinic on eye-tracking analysis of Web sites suggested by attendees, specifically addressing the topic of post-click optimization.  This session preceded several related breakout discussions, including the "Branding and Post-Click Strategies" roundtable session that I moderated. 

Overall, this is a critical topic at a time where more and more marketers are looking for ways to show a higher return on investment for spending in SEO, PPC and other online channels. So in this column I'll cover some of the highlights in the breakout session.

For the panel, a small lab was set up to monitor eye-tracking on three sites, to see if  users could complete a specific task.  While it is worth noting that the testers (conference attendees) were not your average users because of their knowledge of interactive marketing, the exercise illustrated the importance of testing for conversion and search performance.  If there was one single takeaway, it was this: enable this type of testing early in the discovery and design process of Web site and landing page development, testing multiple creative comps, architecture, and messaging on groups that represent targeted users, in addition to testing the pages with a paid search campaign. 


The discussion begins.

There were several immediate points made by the assembled group of agencies, media professionals, client-side interactive marketers and enterprise optimization consultants.  With a user originating from paid search, the consensus was that the most effective types of pages were specifically targeted landing pages (though the pages reviewed by the panel were from various Web sites).  Ultimately the marketer has more control to experiment with conversion performance of multiple pages and messaging.

The conversation then turned toward various challenges of using a site's own pages as paid search landers, as well as the search user experience from an organic result.  While using a site's pages  is often not ideal (in absence of specific landing pages used only for paid search), the reality is that many brand and DR marketers use their site pages as landing pages for very large spends, even though many of these sites were developed without the search user experience and conversion process in mind.  Again, the difference here is in using a Web site page designed generally for multiple purposes, while paid search landing pages are designed specifically for direct conversions coming from paid search.

The paid landing page approach can also become a bit more challenging to scale to catalog sites, or other large-scale publishers.  On top of this issue, even the slightest landing page tweaks on enterprise sites can be problematic in a post-deployment situation.  Marketers are ultimately stuck with a conversion process that doesn't perform as well as it could, and the result is the lower conversion performance of both the landing page and ad spend. 

Most people at the table agreed that getting in front of search and usability issues in the early design and strategy phases of page development will help improve experience and conversions, and increase the return on paid and natural search efforts.  Here are a few other takeaways  from the discussion:

Set up a testing lab, or hire someone to do it for you.   Use focus groups, grab your colleagues or friends, or hire an agency or research group to test before moving forward with a particular concept.  Even if only a small sample can be pulled in for a test, you could still gain new findings to influence design and better inform the conversion approach.

Get media stakeholders involved early in the development process.   If your Web development team or agency is designing a site that will ultimately be used as a landing page for paid media spends, it is imperative to get them involved to anticipate campaign goals and strategies.  The effectiveness of search campaigns often hinge on the effectiveness of the pages themselves.  Investing even a small portion of media dollars in testing and development of multiple comps can increase effectiveness of those spends. 

Test multiple creative comps, architecture, and messaging with your target group, for both direct navigation experience, and also effectiveness in paid search.   Consider varied approaches to testing in terms of messaging and layout.  It is important to do this upfront, so you will have a solid starting point for any tweaks down the road.   Overall, testing multiple concepts provides a better view of what is really going to work.

Use prior performance knowledge to inform redesign or landing pages.  While the post-click session was largely focused on user site activity, our group noted that actual performance data is also crucial to informing future efforts.  Take your performance history from paid or natural search, and ask "What worked?"  Did particular terms convert better than others?  Did the landing pages for higher converting terms have better language or call-to-action than lower performing pages?  Did particular layouts perform better than others? Were there gaps in relevance for your existing landing pages, when compared to your targeted keyword list?  

Build in the ability to test and change.  One of the other challenges of page-tuning is that the process for changing an existing site is often not fluid enough to meet the needs of ongoing optimization.   When you enable the modification of page areas like a heading or opening paragraph, you help provide the post-launch flexibility needed to optimize search campaigns down the road.

Again, these were just a few group discussion points on a major marketing topic that deserves a great deal of consideration. (If you were present at the discussion or panel and would like to add your feedback, please comment on the blog version of this column). As both the panel and roundtable discussion made very clear: Rethinking the conversion process early on can only help maximize and optimize the value of natural and paid search campaigns.