Tracking down information - Search and indexes
A book which has recently landed on my desk is fine, but has a poor index. Hmmm. Indexing is definitely an art but also has some rules, like indexing by surname rather than title. This set me thinking once again about the whole search versus indexing issue on the web.

Many owners of large websites believe that a search facility will enable users to find just what they're looking for. Try tapping 'usability' into the average IT manufacturer website, and you'll get hundreds of similar-looking entries (the example I've always used is the HP website, as I spent many frustrating months a few years ago researching content for HP itself using the company's aimless search engine).

Often, though, the user doesn't know what he or she will find on the site, and is therefore completely unable to specify search terms.

Site search is full of problems. In fact, a report issued by Forrester Research in September 2002 concluded, "Most companies already own a search engine—one that doesn't work."

I'm still a believer in indexes as part of an overall offering for providing access to information, but fewer and fewer people seem to have the skill. I've just posted an old indexing piece in the articles section of this site, but there is other guidance out there too, such as this article entitled Improving Usability with a Website Index on the Boxes and Arrows website.