Loading ...
Sorry, an error occurred while loading the content.

Re: [TaxoCoP] Re: Faceted search analytics - percentage of refinements on search?

Expand Messages
  • Lee Romero
    Hi Stephanie - I would agree with your thoughts here. For the most part, I have found within a solution, the average for metrics like this are pretty stable
    Message 1 of 7 , May 4, 2012
    • 0 Attachment
      Hi Stephanie - I would agree with your thoughts here.

      For the most part, I have found within a solution, the average for
      metrics like this are pretty stable over time.

      It can be useful to understand how your average compares to others to
      give you some idea of where you fall in a spectrum. (As an aside, it
      would be even better to be able to compare your measures with similar
      search solutions and just an overall average but I don't think we have
      enough insight to do that, yet.)

      But the more interesting and actionable things you can do with these
      insights is to compare the measure of individual terms against your
      overall average to identify outliers. You can then analyze details
      around those to try to improve whatever is going on (or, possibly, try
      to improve others if you have individual terms that are performing
      demonstrably better than others for a certain measure).

      The more individual terms you can identify that are problematic in
      some way, the more effective any time you can invest in them is going
      to be.

      Regards
      Lee

      On Wed, May 2, 2012 at 6:30 AM, stephaniealemieux
      <lemieux.stephanie@...> wrote:
      > Thanks to all who replied about faceted search stats.
      >
      > I think the average facet usage rate is pretty low in general, though it depends on the context. I heard a lot of numbers that were under 10% - which almost makes you wonder why we go through so much trouble building the faceted search in the first place!
      >
      > But I think it's also an interesting metric in analyzing search relevancy. High facet usage may be in some cases (more documents/content than e-commerce) tan indication that search results are not relevant enough. Users are compelled to use facets to refine the selection more often because they don't get good results initially.
      >
      > Of course, this is not necessarily all bad, as it means the users have tools at their disposal to make such refinements. But if it's coupled with a high search depth metric (number of search pages viewed), it might indicate that the refinements themselves are not effective.
      >
      > The site I'm assessing right now has an average facet use of 18%, but it climbs to over 40% for some search terms. I'm going to use that as a "canary in the coal mine" in this case.
      >
      > Thanks again to all,
      > Stephanie
      >
    Your message has been successfully submitted and would be delivered to recipients shortly.