I thought this was a validating article for those who have been arguing with me about this for years.
Study: Context Matters for Online Ads
June 16, 2009
-By Lucia Moses
Like print ads, context matters when it comes to the effectiveness of Internet ads, according to new findings from media researcher McPheters & Co. Analyzing the effectiveness of Internet banner ads, McPheters, collaborating with Condé Nast and CBS Vision, found that online ads running on sites with related content were 61 percent more likely to be recalled than ads on sites with unrelated content—a finding that would seem to undercut the case for ad or behavioral networks, which target viewers based on their Internet usage habits.
“The behavior networks are more focused on who they get in front of instead of the environment sites provide,” said Rebecca McPheters, CEO of McPheters & Co. “This says the site on which your ad appears matters.”
The firm also found that not all types of sites are created equal when it comes to recall. Social-networking, shopping and food sites generated the highest recall levels (29 percent to 39 percent), while search and portal sites had the lowest (under 5 percent).
Wouldn’t it be fun to see Google get its ass kicked? Nothing personal, just good fun seeing the big guy knocked down. Well, it’s not happening any time soon, but what goes around eventually comes around.
Check out: www.wolframalpha.com/screencast/introducingwolframalpha.html
Wolframalpha (annoying name, isn’t it?) makes use of contextual algorithms as compared to Google’s general keyword and link-based algorithms—so it’s pretty smart if you ask it something it knows. Wolfram might say about Google, “to a hammer everything is a nail.”
[As a reminder, context is always king—the evolution of the web is about narrowing context to create more and more relevancy and engagement—see the name of this blog if you need proof.]
The first public version of Wolframalpha is more about discovering specific information than it is about discovering websites that contain specific information. Of course, one of Wolframalpha’s subject-matter domains could easily be website discovery—that is, the entire business of website discovery could be just one of the many domain-specific algorithms Wolframalpha offers. On the other hand, what Wolframalpha provides is already a small subset of what Google currently provides, to a degree, so long as the questions posed to mister Google aren’t too tough.
It can be said that Wolframalpha is just the latest flavor of directed search with a few semantic chops on the side. But this time around I believe there is real potential to cover much more domain-specific ground than previous efforts—wikis have taught us much. Can Wolframalpha (or similar efforts) adequately cover what most humans currently discover via traditional keyword search? The current “social search” efforts (let’s leave Twitter et al out of it for the moment) may be irrelevant, but there’s no doubt that capturing the world’s knowledge with domain-specific algorithms will take the work of many hands.
It’s obvious that domain-specific algorithms are more interesting (provide more relevancy) than generic keyword algorithms. Isn’t that why search within Twitter has everyone aflutter? But we’ll all get over the novelty of social discovery and the real-time web soon enough because I believe something much smarter is coming, and on it we will flow.