|Home Articles Effective Search Results||Articles About KB|
Search result pages must make information easy to find and present results in a format that is easy to use.
Most site search engines fall short of both goals.
Searching for "frequensy converter" (misspelling frequency) at Danfoss.com, a worldleading manufacturer, provided no results, while a correctly spelled search resulted in 129 matches. This result violates a basic rule in ecommerce: If users can't find the product, they will not buy. Providing a list of close matches on the search result page would be extremely helpful to users.
Todays cover story in the leading Danish newspaper Politiken is about the crises in the Middle East and mentions the words "Arafat" and "Israel". However, searching for "arafat israel" finds 200+ matches but not the cover story on the day it is published.
Also, slightly misspelling Arafat's name in a search for "arafit israel" returns zero matches.
Despite the known usability advice that users should never see code, URLs are frequently included in search result lists. URLs should be removed from site search results. Codes add confusion to novice users. In particular, complex URLs provide more confusion than help to the user.
Accuracy ratings (e.g., 82% match) shown as text or graphics are seldom relevant. Goal-oriented users don't care about how relevant results are, their only concern is to get the information they need.
Dates should be excluded for the same reasons (except news sites). Knowing that a product page was uploaded Wed, 24 May 2000 10:45:27 GMT is not the kind of information, mainstream users need to order products.
The link to "next" search result page needs to be obvious. However, the "next"-link should automatically remove if only one search result page appears - and jump directly to the content page if only one relevant result appears.
Stating clearly what the user was searching for and how many hits were returned. "Search for ... returned 307 matches". Users can then choose to make a new query if the number of matches is unsatisfactory.
Finally, when slow or overloaded servers result in long response times, outsourcing the search functionality is an obvious decision.