A few weeks ago, Google announced plans to discontinue Google Site Search, a paid product that allows websites to power their internal search engines through Google’s technology. Existing customers can keep using Google Site Search for the life of their current license, but new licenses and renewals will no longer be available come April 1.
In our view, Google Site Search was never an ideal option for e-commerce site search because of its limitations in customizing search display and the relative lack of faceting (filtering) options. It was better suited to content-heavy, non-e-commerce sites that need something very basic. However, the online retailers that were using Google Site Search now have to decide what to do next—migrate to Google’s Custom Search, a similar (and free) tool that adds Google advertisements to search results; build their own solution; leverage their e-commerce platform provider for functionality; or adopt a specialized SaaS service.
For most consumers, finding what they are looking for—quickly, easily and in a personalized manner—is the most important asset for driving conversion. This, combined with the fact that site search is one of the first points of user interaction, makes it a critical feature in the customer journey—one that will heavily determine engagement, conversions and ultimately, profitability. Regardless of the road taken, Google Site Search users (and the broader online retail industry, for that matter) are at a major inflection point. Basic search capabilities have been commoditized, and if more effective search is to serve as a competitive differentiator, what are some important considerations?
Making Search More Intelligent
Most basic site search products are based on keyword matching, which works by simply matching words in search terms to documents and pages containing these words. This type of search is a necessary foundation, but it is often at odds with the wants and desires of modern consumers who now expect speed and convenience above all else.
For instance, consider the example of a user searching for “blue Adidas running shoes.” Keyword-based search will produce all results including one or more of these terms, which could include “blue Adidas T-shirt” or “blue running windbreaker.” In the early days of e-commerce, customers may have welcomed having the widest possible range of results returned, but these days, a deluge of irrelevant results will only risk overwhelming and alienating users, and drive them away.
To meet evolving customer expectations, site search technologies must feature a level of product awareness—an improved ability to dissect search terms and identify what are the key items being sought (in this case it would be sneakers). Any other item or page with matching terms, though NOT sneakers, is either not displayed at all, or pushed way down in the search results.
This greater intelligence and awareness should also apply to facets—the range of options displayed for various products (for example, model, type, brand or function). Facets play a significant role in helping customers narrow, sort, and find products quickly and effortlessly. The challenge is making sure only the most relevant facets are displayed alongside these results (for example, dress may include facets like color, size and brand, but not model or function). If too many and/or irrelevant facets are displayed, this too increases the risk of annoying and alienating customers as a result of a cumbersome, clunky conversion funnel. This is especially true for mobile searches, where too many facets will cause a high friction shopping experience.
Natural Language Processing (NLP)
Today’s retailers are using artificial intelligence (AI) in a variety of innovative ways, and now it is possible to apply AI in the form of NLP to search. The application of NLP to search has traditionally been focused on breaking down conversational sentences. For example, a customer might say, “show me blue running sneakers from Adidas,” and NLP can extract from the term what is actually being sought. The problem with this approach is that it’s not typically how shoppers search on commerce-oriented sites. They use short, non-descriptive terms, in their effort to get into a site, make a purchase and then get out of the site as quickly as possible.
In e-commerce, the far more likely challenge is users searching for the same item, using slightly varied search terms. For example, customers may search for “two cup measuring cup,” “2 cup measuring cup” or “2C measuring cup.” NLP technologies should be able to recognize these searches are seeking the same item, even though slight variations in lexicon are being used. In this way, NLP enables shoppers to use their own language, further contributing to the type of seamless, frictionless interaction that fosters click-throughs, conversions and revenues. NLP technologies also results in search being able to “learn” to identify common synonyms and misspellings, which reduces the likelihood of sites missing conversion opportunities due to simple language nuances or mistakes.
Search Results that Deliver Strategic Advantage
One perceived drawback of Google’s Custom Search product is that it will display Google-generated ads in search results—which retailers will have no control over. While this strategy may be beneficial to Google, it does not help retailers looking to maximize the revenue-generating potential of their sites. In fact, it may increase page abandonment rates as users eyeballs’ drift to Google ads).
Retailers seeking greater control in site search results should consider merchandising techniques supporting “profit-aware” site search, i.e., enabling retailers to define their own search-ranking rules. This helps ensure retailers are always playing their “strongest hand” and maximizing exposure for those products with the highest revenue and conversion potential. Search results can be displayed in accordance with a variety of rules—including newness, brand affinity, click-through rates and conversions. This helps transform site search results into high-performing pages, while also supporting greater marketing and merchandising consistency across the customer journey (international, as well as desktop and mobile).
Personalizing the shopping experience
Customers expectations are evolving at a rapid pace, which includes the expectation of a one-to-one personalized shopping experience, just like a customer would experience in a physical store. Search can play a huge role in this personalization, prioritizing results display in accordance with individual tastes and such attributes as size, color, age, gender, location, brand affinity and style. The goal here is to implicitly understand a shopper’s preferences from previous site behavior, in-store purchase history and other third-party data sources to create a uniquely tailored experience.
Many retailers don’t yet understand the huge power that site search can have in reducing friction in the shopping experience, and increasing revenues and site profitability. They often focus on other areas, like category pages and product pages. The reality is, the effectiveness of site search as a conversion-driver can be dramatically improved, if it is evolved beyond basic keyword matching and best practices are used in site design.
Site search may very well be the most under-optimized e-commerce site feature. Google’s discontinuation of Site Search is actually a timely, strategic opportunity for many retail companies to consider a much-needed site search upgrade. Whichever route they take, forward-thinking retailers will look to harness the immense power of site search, adopting more mature capabilities to help forge stronger engagement, close more sales and increase profit generation.
SearchSpring provides site search and navigation technology to six of the retailers in the Internet Retailer Top 1000.