Scanning Barriers On The Search Results Page
A few columns ago I talked about user behavior on the results page and compared it to a shopping mall. In recent eye tracking studies, we’ve seen that several factors can create barriers or “walls” in this mall that can keep traffic from ever finding your listing. Traffic flows First of all, a quick recap […]
A few columns ago I talked about user behavior on the results page and compared it to a shopping mall. In recent eye tracking studies, we’ve seen that several factors can create barriers or “walls” in this mall that can keep traffic from ever finding your listing.
First of all, a quick recap of traffic flow in the search “mall.” Visitors enter in the upper left, scan down four or five listings, and consider them as a group. Then, depending on the information scent found in the listing group, they either click on a listing, continue scanning elsewhere on the page, or relaunch their search. The important thing to remember is that humans naturally want a few options to consider, so they will make a judgment on the overall relevance of the results set based on the scent found in this initial group. If they see 3 or 4 listings that appear relevant, the overall quality of the results set will be reinforced. If there’s only one or no returns relevant to their intent in this top group, they may likely write off the entire results set and relaunch their search. Your success on the results page is largely dependent on the company you keep.
But the top to bottom, left to right scanning, typically in an F pattern, can be disrupted by “holes” or other visual elements on the page. In this column, I wanted to touch on a few of these.
Distinctions in results
The human eye can quickly determine similarity or variance in the type of listings provided. We may not take the time to read labels, but mentally, we start grouping similar elements on the page into “chunks” for further consideration.
The sequence in which we scan these chunks depends on where they’re located and also where we’ve been conditioned to expect the greatest scent. So, we can pick up that top sponsored results are grouped together, as are the right rail sponsored ads. We can usually determine where the organic listings start without having to spend a lot of time orienting ourselves. We instinctively know our way around a search results page pretty quickly.
As we start to divide the page into these “chunks,” we usually make decisions about the relevance to our intent as a group based on a quick scent scan. We spend most of our time scanning in the areas where we usually find the best results, in the upper left. Therefore, scanning on the right rail or further down the page is less frequent and more cursory. On the right rail especially, we may tend to look at the top listing and if it’s not relevant, we’ll pass judgment on all the ads in this real estate, deciding they’re all not relevant.
We tend to scan “chunks” as groups, and once we’re done with a chunk, even if its scent has been judged with a cursory glance, we usually don’t retrace our steps. We move on to other chunks or launch a new search.
The division of the page, even with subtle design and font variations, can introduce the first set of walls. We subdivide the page into sections — top ads, organic listings, right rail ads — and mentally we build walls separating them. We may pass through the walls once, but once we do this, we don’t tend to go back.
Graphics on the page
Thumbnails, whether they be video or image results, come with one common characteristic. They tend to have straight edges. And the human eye’s need for a sense of order tends to take a straight edge and make a wall out of it.
We extend them out to create mental boundaries to help us navigate the page. Quite often, this will restrict scanning going down beyond the graphic, depending on where it sits in the result set. We saw this quite dramatic effect in the eye tracking tests we ran on Universal results on Google.
Other design elements that can introduce walls are background boxes and thin rules that set off distinct sections of the page. Although these are usually subtle, the human eye picks them up quickly and they act as subconscious guides as we “chunk” out the page.
The Wikipedia Effect
A rather surprising result came out in some recent eye tracking we did. Non-relevant results can act as significant barriers to further scanning, even if they appear in the Golden Triangle. Google’s ongoing love affair with Wikipedia results seems to be a notable example of this effect.
For many many searches, including some commercial ones, Google tends to bring Wikipedia results back in the top organic results. If Wikipedia information is not what you’re looking for, these listings can act as a significant scanning barrier, as users get frustrated and cease scanning further down the page. My feeling is that Google users are getting frustrated with the ubiquity of Wikipedia for so many searches, and the inclusion of a Wikipedia result immediately triggers frustration and curtails scanning below it.
Now, it depends on where the Wikipedia result shows, and, of course, the intent of the user. Look at the example shown, for the keyword “servers”.
If the intent is commercial, e.g., someone looking for product information on servers, a Wikipedia entry is the last thing you want to see. The inclusion of this listing as the number one organic entry would likely signal the death of this particular results page as a search “mall.” Upon seeing the Wikipedia entry, we would pass judgment on the entire organic results set as being non relevant and would likely launch a refined search. If the Wikipedia result were further down, say in the number 3 or 4 organic spot, the impact would be much less damaging to user retention.
The appearance of walls in the search results page can have a dramatic impact on scanning behaviors. If your keywords tend to appear on results pages that are magnets for Universal results, many different types of results, or Wikipedia, you should plan your search real estate strategies accordingly.
Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.