Tuesday, September 14, 2004

Referral logs as intelligence

Twice in the past week my referral logs have told me something I didn’t know. In the first case, it was that a high school classmate and dear friend had passed away. In the second, it was that someone I’d written about a while back is in the news again. In both cases, the spike in number of people searching for those terms indicated an unusual level of interest in those terms; sure enough, with a little searching of my own, I discovered the “new” news.

Which got me to thinking…

If an organization had its employees blogging on a variety of subjects, it could seed Google’s index with a variety of info on subjects across the subject matter spectrum that’s relevant to the company. Then, by running statistics about how frequently certain terms come up (names, subjects, competitors, product names, etc.), they can start to gauge the relative importance of those terms. Importantly, it can be an important leading indicator of a newsworthy event, customer interest, or competitive positioning. An important benchmark would be the deltas between the average number of times a term was searched for and any increase for those terms over a specific period of time.

Cross-referenced against the Internet domains of the people doing the searches, and you could have some very interesting info about what matters to whom, and even start to glean why. Put in the hands of the executives in charge of those business areas, this kind of info could be very useful…


  1. Blog Evidence of DuPage Mass Transit Confusion

    Every once in a while you get a "sign" you're on the right track. Here's what somebody searched for on AOL's search engine today: "what+metra+would+i+take+around+Jefferson+park+to+get++to+Elmhurst" The sad thing is that I know this because my blog is o...

  2. I'm not so sure this would work. Does everything have to be robotic?

    I mean sure you could find the most popular terms and innundate your website with them in hopes for GOOGLE people to find your corporation's site.

    But as a technical writer I would find it quite amusing if you managed to shape your corporation's site after these terms and still retain what your corporation actually is.

    The moment customers find out that your site is nothing but a bunch of 'kick phrases' to garner Google Search People then you'll just get a bunch of one-timers.

    A web site has to not get people to the site but keep them there. Any corporation executives will tell you that they would rather have 500 repeat customers than 10,000 one time customers to their web site.

    Am I wrong?

  3. Instead of waisting one's money on one's employees in an Internet, popular Term Search for Intelligence Information.

    Why not randomly call up customers and survey them?

    I'd much prefer the 'personal' interview or questionairre than what terms that individual searches with on the internet when buying a certain product.

    How would one differentiate between popular search terms that lead to low purchasing versus unpopular search terms with high purchasing?

    I.E. If I searched for 'free of charge whatever service' when looking up a freelance lawyer I may not end up with a lawyer as I would be innundated with links and find nothing of value.
    The word 'FREE' is commonly used term for 'Free Consoltations.'

    But then I search for 'Midnight' and end up with a kick butt lawyer that will work past midnight on any criminal case I hire him for.... And end up hiring him.

    See what I mean? What possible robotic device could you use to tap into someone's server and get the data you need to determine whether a popular term is a good one?