Are Users Ignoring Paid Search Listings?

How much time do you spend thinking about your Fovea? Chances are… not so much. The central part of your retina in which images are projected when we look at something might just tell a story about how we view websites and search results.

Nielsen Norman Group (NN/g) brought foveal action into the spotlight last week with its latest round of usability and eye tracking studies that sought to answer a very important question about user behavior. Are users becoming blind to search listings-- particularly the sponsored listings that marketers pay for and which get displayed in different ways on different search engines?

From the people who brought us the concept of banner blindness, a study in user behavior as part of their Usability Week 2006 NN/ g brought us the Eye Tracking Web Usability day: this was the first instance of a non-search entity sponsored or delivered research, as well as a literal look into the eye of the searcher.

Analysis psychosis
Want to conduct eye tracking research? Strap a camera onto the heads of a few people and watch the fun. NN/g's research was a bit more complicated than that with 232 study participants that were given 50 test tasks in one or two hour sessions.

For the unindoctrinated, an eye chart delivers a graphic image indicating the location of a user's eye fixations during web activity. The image resembles an infrared map of user views with big color coded splotches to indicate where a user's eye was directed most often. The chart is overlaid on the target web site or search result so the researcher can evaluate eye fixation information.
Test subjects ranged from 18 to 64 years of age with 63 percent of users in the 30 to 49 range. They were college educated and 83 percent of them lived in an urban environment. Most had college degrees (69 percent) and they were observed in a "lab" (that is, office) environment.

Users were given open-ended tasks to perform while browsing and searching: like finding a gift for a nephew or locating information on the actual speed at which Mako Sharks swim. They were also given closed-ended tasks such as buying a phone at T-Mobile, buying a shirt at Neiman Marcus or looking up a specific record of baseball great George Brett.

Searchers behaving badly
The combination of qualitative and quantitative data gathering proved interesting, insightful and even comical at times. Qualitative data was delineated as such by the presence of users thinking aloud as they searched or performed tasks.

While qualitative information may decrease the perceived scientific impact of the research (users tended to spoil eye tracking information) the first point to take notice of in observing search activity spoke volumes. There is little time to engage a searcher, and video playback of a search query was so fast that it had to be played back in slow motion.

People still can't spell. Interestingly enough, one of the test subjects spent time in a search result even after recognizing she spelled the query wrong. The test subject viewed the oddly irrelevant results for the misspelled query and later returned to the top of the page to click on the redirect link back to the correctly spelled query.

We have known for some time that misspellings are money makers in paid search. If nothing else, this validates misspellings as an effective search engine advertising or natural search engine optimization practice.

Confirmed knowledge
Users also scanned multiple listings in search results. The click ultimately went to what they deemed reliable, relevant sources of information. Top sponsored links got all of the attention, and researchers were quick to point out that the more closely sponsored results resembled unsponsored or natural listings, the more attention the paid listings received.

Users tended to flip flop from paid to sponsored listings and judged the entire row of sponsored listings based on the perceived relevance or efficacy of the first listing. This behavior should be a warning to advertisers and search sites buying and selling paid listings in bulk without supporting content.

Search is great for newbies and web veterans alike. Another eye chart compared experienced web users with less experienced users. The two eye charts were nearly identical.

Test subjects also tended to read an entire page of results in gathering information before clicking. One user read the entire page of listings when researching the George Brett question.

Not yet blind
If nothing else, eye tracking is an interesting look into the mind of users. The real test of any eye tracking analysis would be changes that occur with site design revamps or year over year analysis. The problem is, with the speed at which search has been changing said time and space analysis probably won't tell us much.

Local search, video search, adding icons to maps, inserting images into directive search results and the constant ebb and flow of contextual listing positions make it difficult to determine the impact of changes to the search or content page with search listings over time.

Though users don't appear to be ignoring search listings yet, the fascination with viewing big red burn-in eye chart graphics on web sites and search results is an interesting look inside the mind of the searcher.