1. Very well put together John. Thanks for the article.

    I’ll be forwarding this article to a couple of project members that are re-examining site search and looking for some best practices. Couldn’t have been more timely.

  2. Great piece.

    I find that a search of a firm’s intranet is just one place to find information. Information tends to be balkanized in different systems with different ways of searching within them.

    I came to the conclusion that there are four types of searching for documents: fetch, recall, research and precedent. You can read more here: http://kmspace.blogspot.com/2007/04/four-types-of-document-searches.html

    I find that these four types seem to be prevalent across the intranet and other internal systems. I think these four equate to your “mode of seeking” section.

    The other interesting item I see is that most people want to browse to all information with an intranet. This may be because most intranet searches do not work very well. Much like the unhelpful example in Figure 2.

  3. …And the unhelpful example 2 may get worse before it gets better. With database-driven learning management systems and knowledge managements systems that allow almost anyone to contribute material, lack of convention is harmful. But whose job is it to establish standards for posting and contributions? These folks might be the ones to offer naming conventions and protocols. Does a law firm assign a steward differently than a corporation with training department? Does the communications team handle this? Or is it up to the IAs and UX designers to moderate such practices? When does the development of a “naming taxonomy” begin to be valued as a strategy rather than a piecemeal approach or a to-do on a project task list?

    The suggestions above are keen and effective…if one has the authority to implement them.

  4. Excellent article.
    One of my favorite search tactics is what I call “browse bypass.”
    If I want to browse, but don’t want to spend time weaving through a sites heirarchy, I’ll search for an item that I know to be within the category that I am searching. Once at the detail page, I’ll latch onto a breadcrumb or some other heirarchical reference to the immediate parent category. Then I’ll back out to the category level and proceed to browse.
    I think that I found this to be particularly helpful when Amazon was in full tab/sprawl mode. I could get so twisted around on their category labelling that it just seemed easier.
    Of course, I’m hoping that the site I’m searching on with provide me with a breadcrumb or other mechanism to latch onto to orient myself once I’m there.

  5. I just got hold of this article… really well arranged. The user dynamics prediction is something to me is highly subjective. and its a good attempt to learn their behaviour. but how would you actually categorize the parameters which can line up and inturn have a good objective clustering done for any user visiting any website. if you could throw some light on those aspects then its the best thing that person like me would be using :)…

    Thanks John

Comments are closed.