Each day we are often presented with an enormous, sometimes overwhelming amount of information online. How do we narrow it down to what we are actually interested in, and what is useful to us? Search can help us do exactly this – to figure out what to do, determine where to go and how to find exactly what we want. So how do we design a search tool within a website that will help us achieve these goals, and one that facilitates what we know about user behavior?
To help answer this, we turn to search patterns. Search patterns are website design patterns for user behavior that facilitate findability and discoverability. Deciding which search pattern to implement into your information architecture should be guided by the expected user behavior, and how you intend to lead a visitor throughout your website.
Let’s take a closer look at some of these patterns:
The most basic, yet highly used search pattern is “exact search” which forms the basis of many more advanced search patterns. It involves basic keyword matching along with stemming algorithm support reducing the results to a base or stemmed form. The search engine is able to index variants (such as plurals, prefixes, and suffixes) of the word and retrieve them, rather than the user needing to enter all variants themselves.
Faceted search allows users to navigate and refine a collection of information by using a number of attributes or facets, reducing the items with those facet values. Faceted search results are grouped using tags which are applied to the page index. The user can continue to narrow down the search results within what they have selected where previous facet values are retained and applied again.
One key benefit from this pattern allows users to take a series of incremental steps down a logical pathway to reach what they are looking for. Faceted search is often utilized by e-commerce and educational resource websites, but can also be prevalent across many other industries.
Parametric search presents multiple options – checkboxes, drop-downs and sliders – to the user at the beginning of their search allowing them to construct the aspects of their query up front.
While this does enable the user to get very granular quickly, it can yield very few or no results due to the number and combination of parameters being searched. This can be frustrating to the user. With the high usage of faceted search, parametric search has started to lose ground.
This pattern is based on the assumption that the user wants to see the most relevant result first. Relevance operates on the principle of showing the user what’s most useful to them first. Ranking based on relevance is constructed from a number of factors which produce a relevance score placing the result higher or lower in the list of search results.
These factors usually include frequency (how often the user’s search term appears in the records’ metadata), exact match over partial match, and the weighting of metadata fields for relative importance. Fields can also be boosted or suppressed (adding points to the score) based on certain criteria that may be not be fully related to the search term.
Pagination helps solve how to display a set of search results that do not fit on one page result.
Some popular patterns include:
scrolls down the page.
The length of your search results, the extent and average volume of search queries, and the comfort level and dexterity of your users can steer you to one pattern over another.
Autocomplete / Autosuggest
Autocomplete will help the user by filling a partial query providing instant page results, while Autosuggest offers related terms for a user. These auto-filled results typically appear after the user has typed an initial amount of characters into a search bar.
Autocomplete is gaining traction in industries across the board, where autosuggest is quite popular with e-commerce sites because of its usefulness for cross-selling, upselling, and promotional purposes.
Implementing one or more of these search patterns can help the users find what they are looking for quickly and easily, providing a more enjoyable experience. The value added is time saved for the user where users will tend to return to a site or app where information is easy to find.