There’s no question. Site search isn’t sexy, but it doesn’t have to be, because the revenue it drives speaks for itself. Internet Retailer Senior Editor Thad Reuter said it best in his February article, Reading Shoppers’ Minds:
“Site search receives little in the way of celebrity-level attention in e-commerce… But with retailers typically reporting two or three times the amount of conversions for site search users, the stakes are obvious: Better site search can translate into more profits.” Period.
However, one question still remains: Which site search approach best connects shoppers with the products they’re most likely to buy, making shopping easier and retailers more profitable? Do online shoppers use natural language search, which interprets subjective terms to serve up search results? Or do the most relevant search results come from learning site search, which “learns” what specific search terms resonate most with consumers and – with SLI Learning SearchTM technology – reranks the order of search results based on the latest activity of users?
In e-commerce, there is some uncertainty around the demand for natural language search. North Face e-commerce manager, Charles Caison, told Internet Retailer, “At North Face, for instance, most shoppers search using terms that describe the product, not ambiguous phrases that require natural language processing to decode. It may be that we have been trained by Internet search engines for keyword searches rather than natural language searches.”
A new SLI study also supports Caison’s insight. To demonstrate site search user behavior today, SLI evaluated natural language terms, focusing on subjective search terms including “cheap,” “nice” and “cute” for a Fortune 100 retailer. As you can see in the chart below, out of 67,000 searches, the word “quality” was only used 3 times while “cheap” and “nice” had similar results. The findings reveal that subjective search terms are not yet commonly used among online shoppers.
|Total searches performed||~67k|
|Searches containing “cheap”||11|
|Searches containing “quality” (high-quality)||3|
|Searches containing “nice”||0|
|Searches containing “cute”||42|
Lakeshore Learning, an IR Top 500 company, also finds less use of natural language search from its shoppers. Lakeshore Vice President of E-commerce Sam Sarullo told Internet Retailer, “an analysis of the retailer’s top 1,000 searches revealed that consumers use an average of 1.8 words to search – a signal that consumers remain wedded to keyword search, and that natural language-type searches may not yet be intuitive. That said, I see return customers who are more familiar with our products using these natural language or long-tail searches.”
The beauty of Learning SearchTM is that if it detects shoppers’ use of longer search terms, it will “learn” and tweak its results to reflect that behavior. Learning Search continuously analyzes the terms and phrases that prove most popular and lead conversions.
Perhaps the best argument for the value of Learning Search is to let e-commerce companies’ results speak for themselves. Here are some of the results leading retailers have experienced using Learning Search:
Some say “sex sells,” but in e-commerce, Learning Search sells more.