By Bill Slawski, on October 17th, 2007At SIGIR 2007, one of the workshops held at the July Conference in Amsterdam was on Web Information Seeking and Interaction.
Web information seeking and interaction involves looking at the way that searchers interact with Web-based content and applications when they are looking for something. The conference covered a wide range of research, and I want to go into a little more detail on a couple of documents that were authored or co-authored by Google Employees.
The papers and working notes from the workshop contain a nice mix of topics, which are worth taking a look at. The papers at that link that initially caught my attention was one on experiments with eye tracking and mouse movements, and another that explored strategies for Web search.
Exploring How Mouse Movements Relate to Eye Movements on Web Search Results Pages
Kerry Rodden (Google) and Xin Fu (University of North Carolina, Chapel Hill)
A mouse click is a proven indicator of a user’s interest in a web search result. In this paper we explore the potential of a more subtle signal: mouse movements. We conducted a study where participants completed a range of tasks using Google, and we tracked both their eye movements and mouse movements.
We discuss the relationship between these movements, and three different types of eye-mouse coordination patterns. We believe that mouse movements have most potential as a way to detect which results page elements the user has considered before deciding where to click.
I’m reminded of Google’s patent application on their Web accelerator program, and its discussion of using mouse tracking to predict which page someone might likely visit next, so that they can preload that page. I wrote about it in Patent Application for Google Web Accelerator?
The second paper that I enjoyed reading was this one from Anne Aula at Google:
Naming the Topic or Reversing Query Terms from Result Documents – Successful Strategies in Web Search
Numerous studies show that the strategies of most of the web searchers are very simple: they use short queries without operators and modifiers or make mistakes with them, and they mostly rely on the first page of results returned by the search engines.
However, the question of whether the users are successful with these simple strategies has received less attention. This paper describes strategies that web searchers have for query formulation and results evaluation and focuses specifically on factors that affect the success of these strategies.
Based on the understanding of the limitations of the users’ strategies, the paper presents ideas on how search engines could more effectively support the users in the information search process by engaging the users in a dialogue-like interaction.
I like that this study doesn’t involve what searchers look for when using a search engine, but rather how they look for what they are searching for.
Do they use special search operators such as plus signs and minus signs in front of some words in their search? Do they refine their queries by adding words or removing words?
How much time do they spend looking at results? How much time do they look at pages that they arrive at from a search engine?
The study also discusses two different kinds of searching strategies that may be worth keeping in mind when thinking about how visitors find your pages if you have a web site:
We have called these two different approaches to querying the straight-to-information approach and encyclopedia approach, respectively. In the encyclopedia approach, the users generalize the terms from the task description (we have also observed this with the users’ own tasks).
These users are using search engines like they were using a paper encyclopedia: they think of a general term that describes the topic and use the search engine as an index for finding sites that are related to the topic. To find the answer to the original question, they browse to the needed information, which is often a tedious process.
On the other hand, the users who employ the straight-to-information approach want to minimize the browsing: they “reverse the query terms from the documents” and aim at finding the answers already in the snippets.
How would you design web pages for your site that can meet both types of search strategies?
One way might to pose questions in your headers. I’m guess straight to info people have longer queries that you might hit on the head by thinking of these questions and using them in your document. Essentially, you’re producing content with specific thought to the long tail of search, which seems to be what is described in the straight-to-info approach, if I’ve understood that correctly.
Questions in headers is one possible approach. I know that you want to create the impression, with what shows up from your site on a search results page, that you provide a quality experience regardless of the style that a searcher might use.
I’m wondering if it makes sense, once your page ranks for terms that you would like it to rank for, to go look at your listing in the search engines and see how satisfied you are with the title/snippet/combination that shows for the result, and if it’s not something that seems particularly persuasive or engaging, to make changes to it and see if the rankings stick while the result (and traffic) improves.
[...] If you could just slap your ad box on the screen like pizza-on-wall you’ve already won half the battle. Eye-tracking always made the challenge of finding where to slap that pizza prohibitively expensive. Because it’s a highly technical, controlled method it cost quite a few greenbacks. There’s a light in the tunnel now: A recent study by a Google engineer and UNC researcher found that mouse-tracking wins hands-over-fist when compared to eye-tracking. [...]
Makes you think that mouse move heatmaps could be even more valuable then eye tracking…
Hi Dani,
The mouse movement heatmaps are interesting sounding. It would be nice to be able to collect that kind of information easily.