Investing in on-site search
One of the questions that some affiliates that operate using a portal mode is the extent of development effort to be invested on on-site search. By its very definition, a portal portrays an aggregation of large amount of content from various sources. However, well organized the site is, users almost always expect a search engine to help them in their decision making process. Which explains the continuing growth of comparison shopping.
If users see a search box on a site (and usually the search option is placed very prominently) they are more than likely to immediately use it. Here’s the problem: search is a reasonably complicated and expensive feature to develop and maintain, if we want to continue to provide users with the most relevant search results. Since free text search is the simplest and most preferred option for users, the site owner’s challenge is predicting how users will search and ensuring the search engine is good enough to throw up the kind of results that users expect.
Are some results better than no results? It is dicey choice to make.
When I have to address this puzzle, I ask myself questions whether a) Is on-site search really necessary for the site being developed? b) If yes, am I willing to invest in keeping track of user search behaviour and using that knowledge to enhance the search engine? c) If the answer to the second question is no, I have two choices. Forget about the on-site search and focus on structuring the site such that the search function is not required (it is a challenge, believe me!) or consider integrating a third-party search engine on to the site (for eg. Google’s search box). The latter option is not without its pitfalls, but one can draw consolation from the fact that the searchers recognize and trust search results from an engine that they are familiar with.
