Improve Search UX with these Three Under-Utilized Optimizations
Site visitors using search typically represent the most profitable traffic segment in e-commerce. According to eConsultancy, up to 30 percent of e-commerce site visitors will use site search, and due to their increased level of purchase intent, typically convert at rates up to five or six times higher than the average non-search visitor. It behooves commerce-oriented sites and apps to get the maximum number of site visitors searching, for as long as possible.
Today, search is going far beyond basic keyword matching to deliver technologies like natural language processing (NLP). NLP enables intelligent search functionality that can recognize synonyms, misspellings and other nuances in human speech, thus improving search relevancy. However, maximizing these newer capabilities – and ultimately bolstering greater profitability through search – also depends on recruiting more searchers and increasing search “stickiness,” and there are several simple techniques to help with this.
Sample Text: Providing sample text in a site search box can be very helpful to visitors. While many sites simply feature text in the box that says “search” or “start typing,” additional instructions and suggestions can guide users to input more descriptors. For example, instead of simply typing “laptop,” sample text can encourage users to specify brand, size or function. This will provoke a search user to search for “Dell laptop” or “HP laptop” instead. Or, before the user hits enter on their “laptop” search, a pull down menu may appear, prompting him or her to display “Dell laptop” or “HP laptop” in the search box. Providing sample text can prompt users to begin using search, while encouraging more descriptors to help derive more context and improve results relevancy.
However, if a site’s search functionality is not smart enough to support these modifiers, you certainly don’t want to instruct users to include keywords that may have the adverse effect of reducing relevancy. Specifically, NLP supports “product awareness,” meaning that if a site visitor searches for “yellow rain boots,” “boots” are intelligently identified as the primary item being sought. So even if a search term includes two of the three keywords, like ‘yellow’ and ‘rain’ - “yellow rain coat” for example – this item will not be displayed, since it is not “boots.”
Persistent Search Queries: Persistent search queries follow search users through the site. After the user inserts a term and hits “search,” the keywords remain in the search box on the next page. This feature is important because search visitors typically go through two or more iterations of a search for the following reasons:
- There are not enough results: After completing a search for “15 inch thin laptop”, the user finds that there are not enough products. They realize they may need to be less specific with their search terminology in order to find the right products.
- Results are not relevant: After completing a search for “15in. laptop”, the user finds that the results do not contain the products they expected. Typically, this is a result of the search engine not understanding their natural language, or not recognizing synonyms. In this case, the engine did not recognize the 15in. is a synonym for 15 inch.
- Too many results: After completing a search for “laptop,” the user finds that there are far too many products - the results may be accurate, but they are too broad for users to find what they want quickly.
Most search users encountering challenges will opt to modify their search query before resorting to advanced features like filtering and faceting. Over 80 percent of these modifications are simply the addition or removal of a single word. Keeping the original search term front-and-center reduces guesswork and enables users to adjust their searches more easily, quickly and effectively. In short, persistent search queries can help keep users searching, while increasing the likelihood of success.
Advanced Filtering: If a site supports searching within specific categories, it is important to let users know. Filters are the key – they can help refine searches on both desktop and mobile sites, although these have different design requirements. On desktop sites, a search scope selector – i.e., when a user enters a search term, this prompts a drop-down menu which allows them to designate a particular category to search (such as appliances, health and beauty or clothing and accessories) – can be very helpful as well as easy to develop. While this may appear to be a feature only useful on sites with tens of thousands of products across dozens of categories, smaller sites can also benefit.
On desktop and tablet devices, filters can simply be placed in a separate column, but this won’t work on phones. Filtering on mobile screens requires extra special consideration, since mobile users tend to rely on fairly broad search terms (given slower typing on mobile phones, compared to desktops and tablets) with the intention of using sorting and filtering options to refine their results.
Mobile sites can provide filtering options in a menu that slides out from the left of the screen, or expands vertically. Ideally, a full-width search bar should be displayed, and the results should be accompanied with a floating full-width “filter options” button. We’ve seen this technique work better than the more traditional “hamburger” menu, which tends to hide most navigation elements from time-pressed mobile users.
Given the immense revenue contribution of search users, e-commerce sites must do everything in their power to attract more search users, reduce any unnecessary search friction and lengthen the search process, while deriving as much context from search terms as possible. Simple search UX optimizations, like those described here, should be considered among the most fundamental search best practices. They are powerful in their own right - small tweaks can go a very long way - but they’re also a critical foundation for more advanced, intelligent search capabilities.
About the Author
Trevor Legwinski is the chief strategy officer at SearchSpring, an eCommerce site search and online merchandising specialist. Trevor is a creative and strategic thinker with more than 10 years in the retail industry, working for Cambria Cove, Strands, and in-store operations for Abercrombie & Fitch. In this time he has also managed e-commerce and marketing for Hallmark, CycleGear and Bambeco.