Why Synonyms Are Critical in Product Site Search
Customer Experience (CX) has many facets when it comes to website search—it requires capabilities like site search relevancy, intelligent search recommendations, and normalization found in Hawksearch Search Information Manager (SIM). Another crucial component of site search is information management. But according to an ecommerce benchmark report, 70% of websites do not offer this capability.
So, what is information management as it pertains to site search? It is the ability to serve up product-type synonyms on your website to ensure relevant results for a search query, such as units of measure.
Most Site Search Tools Don’t Offer Product Type Synonyms
In the study, test subjects were observed to rely heavily on ecommerce search queries that included a theme, feature, relation, or symptom – yet most of the tested sites had poor support for these query types. Even among the 50 benchmarked ecommerce giants, only 34% supported variants in product type searches. And that’s only counting 4 of the 12 essential search query types identified during testing.
In fact, according to the same study, search support was so poor that 31% of all product-finding tasks ended in vain when the test subjects tried using the site search feature. Shockingly, out of the top 50 ecommerce website participants, a whopping 70% of the search engines are unable to return relevant results for Product Type synonyms – requiring users to search using the exact same jargon as the site – while 34% of the sites don’t return useful results when users search for a model number or misspell just a single character in the product title.
Why Product Type Synonyms Can Make or Break Your Site Search
According to a similar study, 61% of sites require their users to search by the identical product type jargon the site uses. In other words, these websites will fail to return all relevant products unless the user uses the exact product term. For example, when the website refers to a “hair dryer” as the product term and the user searches on “blow dryer,” the search will fail to return all relevant products. Ideally, a search for “t-shirt” should yield the same results as “tee shirt” and “all-in-one printer” should yield a “multifunction printer.” From a user’s point of view, these everyday descriptions are as correct as the industry jargon. In fact, most users never think of using a different synonym following poor search results. Instead, they simply assume that the lack of results is showing the site’s entire selection of products.
From the user experience perspective, product type search should have the capability to return relevant results, regardless of whether the searched-for product exists as a category on the site or not. This requires detailed categorization and labeling of products, but also proper handling of synonyms and alternate spellings of those groupings.
Here are several types of synonyms to consider when trying to improve relevancy for Product Type search:
• Near-identical word meanings: ”multifunction printers” vs.” all-in-one printer”
• Regional dialect synonyms: ”spanner” vs. “wrench”
• Regional spelling variations: ”fibre” vs. “fiber”
Other considerations include:
• Text Analysis: The act of “normalizing” text from both a search query and a search result to allow fuzzy matching. For example, a step known as “stemming” can turn many forms of the same word to a more normal or common form to allow all forms to match. For example, taking the words “shopped,” “shopping,” and “shopper” to the more normal form of “shop” allows all forms to match allowing for more variants in search results.
• Query Time Weights and Boosts: Reweighting the importance of various fields based on search requirements. For example, making a title field more important (or weighted higher) than other fields.
• Phrase/Position Matching: boosting the appearance of the entire query or parts of a query as a phrase or based on the position of the words.
• Tags and ontologies: Understanding the query and the document text in terms of specific concepts instead of simply matching terms. Often considered a “concept” search.
• Natural Language Processing (NLP): Understanding the grammatical structure of the text in the query and search results to allow deeper understanding and better matching.
• Statistical Processes: Understanding the relationship between different words statistically. For example, creating code that can detect that “spatula” and “frying eggs” have a level of association.
• Click Tracking: With plenty of data available about user behavior and search, you should leverage it to determine which result is statistically most likely to be the best result for a query.
• Search Engine Plugins: Use plugins that modify the built-in scoring or text analysis behavior to create a custom relevancy algorithm.
• Genetic Algorithms: Use the abundance of data about good search quality to determine the correct values for various weights and boosts that produce the optimal results via a genetic/evolutionary process.
How to Quick Start Your Product Search Capability
Broadly speaking, online shoppers can be split up into two predominant types:
Browsers exhibit a string of behaviors which are the online equivalent to window-shopping. They are shoppers who really don’t know precisely what they are looking for, or perhaps are not sure exactly how to verbally express what they want. Browsers can navigate through multiple merchandise collections, often using the site menu and view many products in one session, without ultimately buying a thing.
Searchers are shoppers who exhibit a clear intent. When navigating a website, particularly an ecommerce website, they are looking for a category of products, a specific product, color, or even a SKU.
Look at Your Top Zero-Result Searches
When we work with companies, we recommend they use their analytics tools to look at their top 25-50 most frequent zero-results searches. A zero-results search without alternatives will likely make the customer think you don’t sell that product, causing them to leave your website (potentially costing you loss of revenue and customer loyalty).
Setting up synonyms for these zero-result searches helps you to overcome this problem. Defining “tv = television” will allow your site search to return “televisions” for both keywords. When synonyms are configured correctly, the search engine ignores capital letters and special characters, applies stemming, and tries to correct for spelling mistakes. So, because there is no need to define synonyms for all of those cases, it’s easier to do than you might think.
Don’t Miss Out on Opportunities—and Revenue
Most ecommerce brands treat on-site search as an afterthought. But don’t make that mistake because ignoring on-site search results typically:
• Lowers the average order value (for desktops)
• Decreases mobile conversion
• Reduces SEO – and thus, less organic traffic
As you can see, the Product Type query is largely a missed opportunity within ecommerce sites and should always be one of the first things considered at the outset and in any search UX improvement project. Otherwise, the likelihood of users getting stuck and ultimately abandoning your website (possibly forever) forever is high.
Ready to Learn More?
Hawksearch is the ONLY unified search, recommendation, and data normalization platform for “complex commerce” that can effectively handle high volumes of Products/SKUs, with complex inconsistent data attributes, and Year-Make-Model and Part Number/SKU searching.
Learn More and see why Hawksearch is the BEST with our easy to use rich features for complex environments, and Hawksearch remains Merchandisers’ and Marketers’ favorite for nearly all types of commerce and content websites and mobile applications.