Laptops generated the largest sales in electronics on Amazon.com in Q1, while wearable technology and gaming accessories are delivering much of the growth, according to ...
Site search can already rank results automatically based on merchandising rules like best-seller status. Next-generation search will sharpen the merchandising focus more by pulling in more analytic data.
Better-that is, more relevant and therefore more likely to prompt a purchase-site search results are key to better merchandising at online stores, and the next generation of site search technology is aiming in that direction. “The next frontier for site search is combining search and navigation with analytics and merchandising rules to create a self-tuning, automated system that constantly improves merchandising through the search results it delivers,” says Michael Crandell, CEO of search and navigation technology provider Celebros Inc.
That combination will enable site search that ranks the results it delivers according to criteria set by retailers, for example: listing the best-selling items first or relevant items of which the retailer has an excess of inventory and wants to move. Qwiser, Celebros’s search product, already can rank results according to best-seller status, but future iterations of the product scheduled for release this summer and beyond will incorporate even more analytic data into the ranking of search results, Crandell says. Crandell notes that while the product is compatible with the web analytics packages of outside vendors, it also incorporates its own analytics tools that use natural language processing, like the search product itself does.
“In the future we’ll be combining a broader range of analytics tracking to create a self-tuning merchandising system for online stores. There is a lot more to merchandising than sales ranking, in terms of tracking particular paths that customers use and their specific behavior on a site,” he says. That broader tracking will eventually feed analytic data on a customer’s behavior, as well as demographic and geographic factors, into Qwiser to rank search results. And that would make results on the same site search query as well as the choices on search-narrowing options different for each customer segment, even for different individuals, Crandell says.
“It will depend on what technologies, such as geographic locating technology, the store is using. We want to be able to build the tracking into the search process,” says Crandell. “At the end of the day, what we are trying to do it get at what the user has in his head.”