Polaris, the new site search tool recently launched by Wal-Mart Stores Inc., is already generating extra revenue after just three months.
The revamped algorithm for Walmart.com’s site search engine displays results based on three measures of product popularity: most recently purchased, most posted about on Facebook and most positive reviews. Items which other customers have recently searched for or clicked on are also displayed.
Smart searching
The new search engine, which was built over ten months by just 15 engineers in @WalMartLabs (the company’s team of mobile and e‑commerce platform staff), provides plenty of food for thought for the innovative e‑commerce analyst and e‑commerce manager alike. The Labs’ vice president, Sri Subramaniam, said that it understands customer intent much better than its predecessor, Endeca, adding, “retailers embracing e‑commerce have to very quickly turn into tech companies — and search is the crown jewel of this.”
In an interview with the AllThingsD.com, Subramaniam revealed that, since its implementation just three months ago, Wal-Mart’s online sales have surged by 10 to 15 percent.
Departing from the conventional search technology used by other bricks-and-mortar stores for their e‑commerce wings, Polaris has taken a leaf out of purely online retailers’ design books. E‑commerce entities like EBay and Amazon use search technology that constantly refines the search experience. Just think of searching for a book on Amazon, as soon as it’s displayed, customers see: “People who bought this also bought X”, where X refers to one or more related products. The combination of search accuracy and offering a range of related products has proved highly successful commercially.
It’s all in the semantics
The new development reflects the fact that several of the engineers responsible for Polaris came from the semantics technology firm Kosmix Corp, which Wal-Mart acquired last year. Semantics technology focuses on what consumers mean when they enter search terms or make comments online.
As Subramanian explains, the older, less semantically sophisticated search engine was likely to display dog houses when consumers entered the word “house.” The new search engine brings up DVDs for the hit TV show “House” at the top of the list, which turns out to be what most people are actually looking for. It also displays subcategories so customers can browse through related products. As you can see, smart searching just came to Wal-Mart.