Site Search Best Practices: 5 Ways to Test Your Functionality

For a quick reminder of how important site search tools are to e-commerce businesses, consider the following statistic. On the average shopping website, about 1.8% of all visitors will make a purchase; visitors who have used the search tool on that same site will convert at a rate of 5.2%. 

In other words, users who run a search on a given website are almost 3X more likely to buy a product or service.

Clearly, optimizing the site search functionality is critical to e-commerce success. So we asked our friends at Klevu to help us put together a list of site search best practices. Read on for recommendations from Klevu, including both the features a strong search tool should have and the simple ways that you and your team can pressure-test your current functionality.

 

Start with a question: How visible is the search functionality across my site?

Given the positive impact that site searches have on website conversion rate, a crucial best practice for e-commerce platforms is giving the search bar plenty of real estate and emphasis. The header is an especially important place to put search front and center, so that shoppers can find the search bar easily on any page.

Additionally, Klevu recommends further upping the visibility of the search bar by adding pre-filled text that makes clear to website visitors that they can use that form to navigate to any product or service they want.

 

Then test your site search tool for best practices.

Klevu shared a number of must-have features for strong search tools, and those are listed below. Klevu also noted that there’s a very simple way to check if a given website has these features in place: just run a bunch of different searches and assess the quality of the results you get.

Here are some of the most important features to test for:

 

Autocomplete suggestions for search terms

To help customers get from the search form to the products they’re looking for as quickly as possible, autocomplete suggestions are a must. Klevu calls this “search-as-you-type.” For example, when a user on an apparel website starts their inquiry with “dress for…” the search tool should suggest popular words to complete the term: “dress for weddings,” “dress for summer,” etc.

 

Try this out for yourself to check whether your site search uses predictive technology. If you type in “tee…” does it suggest “tee-shirt”? Or maybe even “tea dress”? If you’re not seeing any autocomplete suggestions during the test, your functionality has room for improvement.

 

Recognition of misspelled words and other errors

Search leader Google has shared that “every day, one out of 10 search queries is misspelled.” Plus, “new words are constantly being introduced, along with new ways to misspell them.” In other words, it’s safe to assume that around 10% of the terms customers search for on any given website are misspelled. It’s essential to the e-commerce user experience (UX) that site search tools can recognize errors and return the correct results.

To test your search tool’s ability to tolerate errors, try a variety of misspellings. From a simple letter-swap to total gibberish, different styles of errors should all be handled in a seamless way by your site search.

 

Understanding of common search term synonyms

A strong search tool will have what Klevu calls “wide coverage” of related words, and be able to discern the customer’s intention even when they use synonyms for product features. For example, someone searching for “navy couch” should get quite similar results to someone who searches for “blue sofa.” This requires the addition of contextually relevant synonyms to product catalogs (which Klevu does automatically!).

For this feature test, think of a popular product from your catalog. What are all the different ways customers might search for this item? Enter as many terms as you can think of, and see whether all those inquiries eventually bring you to the same page. 

 

Voice search capabilities

This last feature is more of a “nice to have” at this point in time, but brands looking to stay ahead of consumer expectations should absolutely be thinking about voice search functionality. Coinciding with the rising popularity of voice-controlled devices like Amazon Alexa, shoppers are getting used to vocalizing their product searches: Google says that 20% of searches in its app are done by voice, and this number is expected to increase over time.

You probably know whether your site has voice search functionality already—but you can still do a quick test of its sophistication. First, say a simple term, just one or two words. If that pulls up the correct results, that’s a good start. Next, try a search term of four or more words, like “black party dress under $100.” To decipher a complex phrase like that one requires Natural Language Processing (NLP), a feature that will become ever more important to e-commerce success as consumers get more and more comfortable with voice search. 

If you’re wondering how well these site search best practices can work, look no further than the company that dominates e-commerce in the US: Amazon. The Amazon.com homepage is massively oriented towards search, because its leaders know that when users search for a product, they’re more likely to convert.

Ready to level up your site search even further? Note that Amazon also features popular products on their homepage based on what people are buying and user behavior. This is an AI- and ML-powered feature that anyone can access by partnering with Klevu.