Improving Search Usability

We need better tools for refining/filtering search results.

37 Signals has touched on this before:

Yet as good as their search engine is, the results are only as good as the search terms you enter. How can you know if “african coffee trade” is a better search than “africa’s coffee trading” when you’re doing research? You can’t possibly know, but Google can.

Currently, the most obvious method of refining a search is through navigational tabs. Google, Yahoo, MSN, Ask Jeeves all have a similar set of tabs (web, images, news, local, and products) and a more extensive filter set available through advanced search options.

The problem with search engine defined filtering/tabs is the restricted limit. In order to present more filtering options, search engines need to place more tabs on their search pages – eventually they will run out of screen space.

In the past, Danny Sullivan of Search Engine Watch has suggested invisible tabs as the solution for refining search results. Search Engines work behind the scenes (ala Google Image results or Ask Jeeves pictures) and return different sets of data depending on the user’s search.

This works for predefined data types (images, videos, local, news) but does not allow for you to refine your own search. If a search query doesn’t return what you were looking for – you need to input a brand new search.

Sure you can modify your search terms and get a better set of results, but why should you have to? Why can’t search engines provide an interface so we can filter results based on our search query?

Personalized search is one step towards providing a better tool for search filtering.

A9′s Open Search and Google’s Personalized Search are both testing new personalized search technologies.

Google delivers results based on user profiles:

For example, people with an interest in the outdoors will see different relevant sites for a search on “bass” than people who are interested in music.

Open Search allows for filter searchs through other content providers (vertical meta search):

OpenSearch is not a search engine—it is a way for search engines to publish their search results in a standard and accessible format.

Before I can use any type of personal search, I need to define my profile – in Google I pick my interests and in Open Search I select the content providers. Either way I am actively filtering my results before searching. But what if I don’t have the experience or knowledge to filter my search beforehand?

Let’s say I’m looking for information on how to center my webpage.

My first search at google for “center web page” brings up lots of results with the noun form of “center”: Space Center, Kenedy Center, etc. Now if I had typed in “centering web page” I would have been taken to the results that I wanted.

Why can’t search engines provide me with the tool to refine my search AFTER the initial query?

Google understands enough about the word “center” to know that I could be searching for a multitude of different items. Yet I get no method of altering my search.

Here’s how search engines could improve:

  • Contextually Defined Tabs – Provide tabs that I can click on that will refine my search based on my initial search query. If I’m searching for “SEO Forums”, I doubt I’ll need the froogle tab. But a tab that narrows my search to “Search Engine Optimization” websites would help.
  • Clickable Search Alternatives – What search queries are related to and stem from my original search. Please make it clickable.
  • Keyword Tags – Display the relation between web page results and the keywords used to find them. If most people are finding/clicking-through to CNN after a search for “news”, then tag the CNN page with “news”. This way I have a power search tool. If I want to find ipod sites with news (aka ipodlounge), I’ll search for “ipod news”. If I want to search for ipods on “news” sites (cnn’s latest news on ipods) – I can do that.