Calendar
Network
Feeds
- 56
- Sphinn It!
Went Hot: April 13, 2008 - 7:06 am
Posted By: jeffquipp 450 days ago
Topic Type: News Story (Jump to http://searchengineland.com)
Category: SEM Industry
Good information to have when constructing new content for client sites!
4 Comments
Comments
The paper is available on Dr. Jansen's Web page (http://ist.psu.edu/faculty_pages/jjansen/) at Penn State's web site, and was published last year:
Determining the Informational, Navigational, and Transactional intent of Web Queries.
http://ist.psu.edu/faculty_pages/jjansen/academic/pubs/jansen_user_intent.pdf
Data for the study appears to have come from Dogpile from 2005.
Dan Russell, from Google, gave us a different and more detailed breakdown of query types from Google searches in a November 2006 presentation at Stanford. See: http://searchengineland.com/061218-231343.php
Thanks Bill for links to the full paper. My only concern with the data is in my own studies I've found that navigational searches make up 5-7x what they cite in this paper.
I'm attributing their data collection sources - dogpile and infospace - as the reason for these percentages. Google has a nearly ubiquitous toolbar and is the "navigational gateway" for so many who bypass the browser address bar, so I have to think a new study using the major SE's search logs, while improbable, would yield far different results.
I would support the notion that the 'true numbers' would show much more navigational searching based on what I see in referral logs. Perhaps Google with its 'I'm feeling lucky' button was on to something.
Does this look at each search as an independent event or does it look at the string of searches a particular user may have done in one browsing session? We have found using our tracking technology that users that have intention to buy may start with what looks like an 'informational' search, then get more granular with a 'transactional' search, and then finish off with 'navigational' search to get back to the site that they want to perform the transaction on. When we look at this search pattern as a string vs. 3 independent searches it tells a different story.