Mobile accounted for 25% of Ulta's e-commerce revenue during Q2.
Shoppers speak through the site search box, and e-retailers can profit by listening closely.
Site search doesn't lie, says Jason Miller, Motorcycle Superstore Inc.'s chief technology officer and vice president of technology. That is, site search data provides retailers with an unfiltered look at what consumers are actually seeking and the ways in which they're trying to find products they're interested in buying.
Because site search data offers a window into the consumer mindset, the online-only retailer Motorcycle Superstore pays close attention to on-site searches—and even occasionally abandons the conventional search results page to move consumers closer to the Buy button for popular items. The result has been higher conversion rates, says Miller.
The retailer of motorcycle gear digs into the terms shoppers are typing into the e-commerce site's SLI Systems-powered search tool, hoping to discover the terms that consumers actually use when shopping—including common misspellings—seeking to add to the phrases manufacturers provide or that the retailer's own staffers think shoppers might use. When the retailer discovers a trend, such as dozens of consumers searching on the same unexpected term, the retailer often adjusts its paid search marketing campaign to bid on, or increase its bids on, those keywords, Miller says.
Miller isn't alone in employing site search data as a tool for driving online sales. 66% of businesses in a recent Forrester Research Inc. survey said they plan to expand their implementation and use of site search data this year. But the hard part often lies in how to use that data, which can require manual intervention to make sure a retailer is displaying the products and information consumers want.
For instance, Motorcycle Superstore custom-builds landing pages that serve as de facto category pages for commonly searched terms and take the place of the standard search results page.
When a consumer searches for a "Shoei helmet," products produced by a major helmet manufacturer, he arrives at a landing page featuring a banner image that displays the Shoei logo, two helmets and a message promoting "Fast, free shipping." Below that central image are other images featuring five styles of helmets, such as "full-face Shoei helmets," "half-helmets" and "flip-up modular helmets." Below that are links to enable a consumer looking for a helmet for a specific use, such as dirt bike motocross, to search by activity.
Those custom-built category pages, which can take anywhere from 15 minutes to a couple of days to create, give consumers a more coherent view of the retailer's offerings than they would receive just from search results, says Miller. "Rather than return a search page, we're automatically putting the shopper on a category page," says Miller. "That cuts out an entire step."
The category pages are built using the retailer's in-house content management system. The system enables a member of the retailer's e-commerce team to set criteria that control which items are included on the page. Or the page author can use widgets to drag and drop items onto the page. Either way the page is then dynamically populated with the correct data. Occasionally the process takes longer, says Miller, because the retailer is building a category page that requires research to craft original copy or to build a variety of menus to help guide the consumer to a specific product.
Motorcycle Superstore has created between 800 and 1,000 such customized landing pages after determining that the conversion rates for the customized pages are 4% to 5% higher than for conventional search results pages, says Miller. "The conversion rate goes up because the pages are focused on what the consumer is searching for," he says. "It enables the customer to get into the department pages quicker." That produces substantial results, he says, because the fewer clicks it takes a shopper to find what he's looking for, the more likely it is that that shopper will convert.
The goal of improving its conversion rate is what drove cleaning and facility supplies retailer BettyMills.com to rethink its site navigation. The retailer discovered that consumers using its site search had a 100% better conversion rate than other site visitors. That meant that consumers browsing the site just using navigation tabs weren't finding what they were looking for, says Victor Hanna, the company's CEO, whose official title is Chief Betty.
Because the rest of the site was built in-house, fixing the problem would have required the retailer to manually revamp the product organization on its category pages. Instead the retailer decided to use Celebros' search results to populate those pages. "We figured that if search produces a better result, why not have every user effectively perform a search when they navigate the site?" The result was a 212% year-over-year increase in time on site in May, a 100% jump in conversion rate, and a 10% bump in average order value. "That is dramatic—those aren't small gains, those are huge gains," says Hanna.
Routing everything through search required Celebros to make some changes to ensure that the newly created pages would show up high in search engine results, he says. That included making the URLs for those pages intelligible to search engines. For example, when a consumer clicks on the link to the "brooms" category page the page's URL is "www.bettymills.com/shop/product/find/Brooms" rather than a long string of seemingly random symbols and letters. Search engines give more credence to pages whose URLs contain the terms that online shoppers use when they search.
With new products regularly coming in, reworking the search results to ensure that consumers are finding what they're looking for is an ongoing project for the retailer. "The number of moving parts is extraordinary," says Hanna. "We're constantly making changes to our inventory so we have to be on watch."
The Betty Mills Co. doesn't leave that task to one employee. All staff members are charged with alerting the merchandising team if a search produces an odd or unsatisfactory result, for instance, if a mop appeared at the top of a search for a "corn fill broom." If that happens the merchandising team tweaks the tagging on applicable products to improve results.