Much has been written about the correct placement and style for effectively integrating search on a site, yet a fundamental issue with such search remains: the results are too broad and are difficult to sift through. Alleviating this problem is a simple case of letting the user put their search query into context. That is, if we provide a mechanism for letting the user search within specific categories, the probability of that user finding information pertaining to their interests is greatly increased. This mechanism is what I refer to as “categorical search.”
Categorical search is by no means a new idea; in fact, it has been around for hundreds of years. Can you imagine searching for a book in a library or a video in a video store without genres? What if you had to search through an Atlas for a city without being able to choose the country first? How useful would the yellow pages be if business types were not grouped together?
Although categorical search still exists, it is greatly under-used. The problem is that we assume that new technologies can replace this tried-and-true tool by creating smarter algorithms. Yes, our algorithms are getting better, but they are not perfect. By adding another variable to the calculation, accuracy can be heightened and the number of irrelevant results reduced. By allowing the user to select areas to search, the search tool can perform a more effective search, while improving the user experience by returning results in line with that for which they are looking.
The most common implementation of the categorical search on the web is a drop-down menu within the search module containing categorical options for setting the context of the search as shown below:
Categorical search example from ebay.com
The above example presents the easily recognized search text box and search button with the categorical drop-down list between them. This layout ensures that experienced and novice web users alike recognize the ability to select a category in which to find their results.
Since experienced users generally type into a search box without hesitation, a drop-down preceding the field may be lost. As such, placing the categories in between the recognized search box-button pair increases its visibility for the web-guru. For the novice user, who generally takes more time to review the information, reading as they would on paper (left-to-right), this layout is also very well suited as it creates a logical relationship that connects the items in a in a familiar way. An even better indicator for the novice user would be to precede all options by the word “in” to explicitly declare that the keyword will be found within the listed categories.
Another important aspect of this layout is that the default value of the drop-down is a valid one, allowing users who prefer to skip-over this selection to receive the default experience of searching all available categories.
The former example was taken from an e-commerce site which is currently the type of site most prone to implement categorical searches. A common assumption is that this type of search is only really suited for e-commerce sites, where users are searching for a particular product. However, contextual search is applicable to every web site where the user seeks specific information, which is to say, pretty much every web site on the World Wide Web. The rest of this article will explore the application of contextual search within:
- E-commerce Web Sites
- Content-rich / Informational Web Sites
- Corporate Web Site
- Community Web Sites
E-commerce Web Sites
E-commerce web sites have been the widest adopters of categorical searches because the mechanism lends itself very well to the process of finding products.
Let us think of the way a user generally shops in the real world. There are those that take a list for specific items that they need (either written or in memory) and those that go to simply browse around to see what they can find.
In the case of the list-maker in the supermarket, he/she uses the aisle markers to find what they are looking for: apples and lettuce in the “fruits and vegetables” aisle, cheese in the “dairy” aisle etc. For the “browser,” he/she may go from aisle to aisle looking for what they want, or simply choose the aisles of interest. In both cases, the user can use categories to aid them in finding something useful. It is unlikely that either would want to peruse the “baby products” section if they don’t have a baby for which to shop. This ideal was easily transferred to web, hoping to improve the search experience and expedite the buying process.
E-commerce sites provided a mechanism for both browsers and list-directed users to narrow down the products by type and thus increase the likelihood of finding something to buy.
Content-Rich / Informational Web Sites
For content-rich web sites the advantage of contextual search is also self-evident. In the same way that libraries have been categorized because of the wealth and depth of the information it offers, the informational site needs to allow filtering by topics to provide more relevant results.
For example, if a user visited Wikipedia and wanted to find the entry regarding “Michael Jordan,” key researcher in Artificial Intelligence, that user would know to select to only search for results that match in the “Technology” category, in order to avoid the need to sift through the results matching the more re-known Michael Jordan of the “Sports” category. This is a prime case in which categorical search improves the user experience while reducing the weight on the search tool.
A rising informational site set that provides tagging and categorical browsing but oftentimes neglects such filtering within search, is that of the Blog. For those that have a great number of posts, this is a very handy tool for letting users find a specific post within the tags of their choosing (don’t worry this blog will benefit from such as our content expands).
Corporate Web Sites
This is the type of site most resistant to the idea of categorical search; however, it is the one which is also moving fastest to needing this feature. As companies expand their offerings and as more and more of them introduce articles publications, case studies and other in-depth materials for the competitive edge, categorical search can serve a great service of targeting users to the section which appeals to them most.
For companies that offer products and services categorical search is particularly important, as users generally know which most interests them. They may be looking for a specific product without knowing the name, and failing to find it via browsing, turn to the site search. If that search returns primarily services on the first page of results, this can frustrate the user and cause them to move on. The point of cross-selling services and products is a strong one which may be made, but I would argue that search results may not be the ideal place to cross-sell, particularly if it will present an impediment to users’ goals that may drive them away from the site.
Community Web Sites
Community websites are a great place for categorical search as there are distinct areas of interest and activity: blogs, friends, groups, events, discussions/forums, and (of course) media. Due to the fact that community sites are mainly exploratory, which is to say, users move through the site through links scattered across the site and the goal is to find by perusing, when searches are done, they are generally targeted at a very specific section. For example, looking for a username has to be an exclusive search so as not to show items of the same name in blogs or the like. In many cases, this categorical search is presented through context, and thus if a user searches while viewing groups, only groups results are returned. This mechanism is very effective once the context is explicitly made available.
Overall, web sites are increasingly being designed to be “user-friendly,” with particular emphasis on the site architecture: categorizing information so that they can be easily found. By employing the categorical search, we may leverage this work to also increase the usability of search so users can both browse and search with satisfactory results. More discussion will follow about how to categorize information and maximize the effectiveness of this approach.