Just a quick heads up (and reminder to myself to post more on this idea) about a new Yahoo search trends service – Yahoo Clues (official blog announcement). As well as allowing you to check search term trends over the last day, week or month, you can also get information about the demographics of the searchers (gender, age range), as well as their location. The video below suggests that the search volume and rough location data is associated with government data on income by location to guesstimate income distributions for searchers. At the moment the service is limited to the US, but the page design looks hopeful as far as opening it up to other territories goes. That said, the geographic filter doesn’t seem to apply to the displayed demographic data (yet?), so it currently looks as if you can’t compare demographics of searchers in one area with those in another?
Trends are shareable via a URL, although I don’t think you can specify a date range (yet?), so if you see an interesting chart, you probably should grab a screenshot rather than trust to being able to revisit that particular view at a later date.
I don’t think the pages are live, either, in the sense that if you view a trends results for the last day and leave the page open in your browser, I don’t think the data automatically refreshes?
So – who’s searching for “library” in the US? (Clues about ‘library’):
Ah, right – the 25-34 year olds are looking for iTunes library information;-)
But the younger ones want Proquest?!
The “search flow” tool appears to borrow from a related idea in many website analytics pages, where you can view the previous page visited and the next page visited relative to a given page. The example I’ve chosen (‘library’), is not hugely informative, though if I was to clutch at straws I could maybe try to argue (weakly), that the tool shows refinement (going from ‘library’ to ‘library of congress’, say, to generalisation (going from ‘library of congress’ to ‘library’)? Err…?!;-)
I’m not sure if Google Analytics allows this sort of tracking on searches carried out on tracked websites, but the “search flow” of students using the Library website for a particular research task might well be instructive in terms of gauging the level of users’ search skills, as well as the strategies they appear to be using? For instance, it might reveal cues (keywords, key phrases) that students pick up on, as well as the way they refine searches, maybe in light of search results achieved from previous searches?