Steve Rubel at Micropersuasion conducted a nice analysis on Wikipedia influence in search results:
Wikipedia articles on the top 100 advertisers in the U.S. are consistently among the most highly ranked pages in Google on direct searches. This is according to an exclusive Micro Persuasio study conducted over the past week. The study was compiled by simply taking the largest 100 advertisers from AdAge, entering them into Google and then tallying the results.
The obvious conclusion here is that most search results are full of related entries from Wikipedia (not to mention blogs and other peer media). Taking this a step further, this means that millions of individuals who hear about products through these billions of dollars advertising and then turn to Google to research them are likely influenced by what Wikipedia says. Further, many searchers probably do not even know that the encyclopedia is not run by a single body but by a collective of individuals, even people they may know.
Steve’s work is interesting, but his results don’t necessarily prove his conclusion, and the underlying story is far more complex. My colleague Mathew Hurst pointed me to Graywolf, who analyzed AOL search data and countered:
Now you don’t need a fancy college edumacation to tell you the higher up the result the more likely it is to be clicked, so lets grab the brands with the highest wikipedia rankings and look at them compared to the AOL data. Coming in at #3 we have Directv, Circuit City and Acura.
Lets what the people who actually went to wikipedia were searching for … hmm well it seems to center around something that stars with an “s” and ends with an “x”.
Now all of this data could simply be a byproduct of the less sophisticated AOL user base, or it could be something else
To Steve’s point, Wikipedia matters in its influence of search result shelf space, if anything. A search result is a brand impression, and there are varying levels of incremental impact, even if not clicked on.
But despite Graywolf’s findings that Wikipedia is generally not clicked on that much, it still is clicked on at a much higher rate versus other sites, after visiting a search engine. When I was working with Hitwise on their search behavioral data over a year ago, I asked that question: How does Wikipedia rank among all sites based on dowstream click-throughs from all search engines? Here was my key finding, which I wrote about in MediaPost:
With nearly 600,000 "living" articles to date, Wikipedia’s orderly collection of consumer-created content is becoming a high-powered magnet for Internet searches. A ranking of all Web sites based on the total volume of traffic received directly from search engines placed Wikipedia at 146 in June 2004. But in September 2004 it jumped in the ranking to 93; 71 in December 2004; and in March 2005, it was the 33rd most popular site in terms of visits received from search engines.
The fact is that relative to all sites, Wikipedia does get significant click-through, though that’s against a very long-tail average. Search shelf space doesn’t imply the sky is falling, it only implies that there is some degree of potential for the sky to fall. The key is to understand which search results are occuring, and which are actually being clicked on, by whom. This is where Web audience and search analytics intersect with public relations. There are not too many firms who’ve mastered all three, or really strive to. Converseon does some interesting work in this area.
I’m talking with Hitwise about re-running these data for me.