Search Engine Users

The PEW Internet & American Life Project released its “Search Engine Users” Report January 23rd, 2005. I’ve printed it and given it a quick read today.

Interesting:

  • The average user spends about 43 minutes a month doing 34 searches, viewing 2 pages per search (I search way more than that)
  • 92% of searchers feel “confident” in their searching abilities. Only 27% of them actually find what they’re looking for every time. Is that a mistake of finding something, not the thing?
  • 44% of searchers stick to a single search engine. Only 7% use more than 3 search engines – they tend to be the most confident of finding things, take their searching more importantly than others, and search multiple engines deliberately, knowing some engines are better at different tasks than others (I’m squarely in this group).
  • That single-engine 44% sticks with (usually Google) because it’s usable to them, without them necessarily knowing so. Stated reasons include speed/accuracy/comprehensive results. They don’t know better. Or they might suspect, but don’t care.
  • “92% of users demonstrated that if they were unsatisfied with results from a search, they would rather launch another search at the same engine than switch to another search engine[.]” This suggests a willingness to accept responsibility for not finding something at an engine due to “not asking the question correctly” instead of considering first, “the answer may not be here”. Trust or Laziness?
  • 68% of users think search engines are “fair and unbiased”. 60% of users (in a different study?) didn’t know some listings in search results are paid ads. Only 1 in 6 users can tell an ad from a result. 54% of those users who recognize ads click on them anyway. So go ahead and advertise in the biggest search engines. The users just assume it’s really what they’re looking for.

Thinking Differently About Site Mapping and Navigation

Keith Robinson smacks the nail squarely on the head, IMHO in Thinking Differently About Site Mapping and Navigation, a topic that’s been bugging me lately, as a large site redesign looms.

Since the content for the redesign isn’t here yet, I’ve been “fantasizing” about how it would be best offered to users.

My favorite bits, along with my own thoughts plugged in:

  • “Home” isn’t where the “start” is. In the age of Google and RSS, most people probably arrive at a web site several layers deep. Navigation, Presentation and Related Content cannot fail to be there, or they go right to their “Back” button and find another site in their search results to look at, perhaps never to return.
  • Sitemaps tend to be hierarchical presentations of a site’s page topics according to the author(s), and cater very poorly to the finding of content a visitor is searching for (unless the page title really does encompass the whole page’s content). Useful for designing a site among content creators, not so much for content searchers.
  • “Home” might be more useful as a “Hub” (ironically, a site map), because who hits a homepage except: Someone who meant to come to the site, Someone who’s already read something and wants to see the site from the “beginning”; or Someone who comes back often, because the content they’re looking for is easy to find from there.
  • “Indexes” are more useful than hierarchical “sitemaps”. Yes, Indexes: “A… B… C…” etc. To be truly useful, they ought to be dynamically created by the same system that provides a site-based search, so they’re automatic and all encompassing. Or, limited to the public areas of a site, like a site-based search would be.
  • Remember: the best “help” files I’ve ever used have an index, a content search, and related items accessible from every page. Very rarely do I actually use the first-shown tab of the help file: the tree-based “Contents”. What else is a web site, than a “help” file for information I don’t already have?

A related thought: “Folksonomies” sounds like something George W. Bush would say about a foreign country, but in reality it’s how the masses each uniquely (and thus, collectively) describe content in just a few summary words, each. See Del.icio.us for a great example.

Ideally then, what I want for a web site is a content management system that offers search, showcases the total range of topics, the latest topics, popular topics, etc. as a part of the “home” page, and collects/tracks/offers folksonomies, keyword density, user comments, questions-and-answers, related items and a context-relevant index as side-bar items on every page. It would, in theory, make the site index, site search, and related items parts of a site “automatic” and relevant based on what the people who see the site think of it.

If only they’d contribute, to a site meant mostly as a description of services, facilities and locations, not a collaborative discussion of topics.