Advances in Online Retailing Search Technology May Turn Browsers into Searchers

When it comes to finding products on an online retailer’s website, a customer’s shopping method can likely be classified as being either a browser or a searcher. Browsers are best thought of as those that like to locate products by entering a site and then taking time to look around by clicking down many levels of a product category. For instance, if someone goes to the Target website looking to purchase a toaster, they would start their shopping by clicking on the general Home, Furniture & Patio top level category then follow a click-path that takes them through the Kitchen->Kitchen Appliances subcategory and then into the Toaster sub-subcategory. Depending on how well a retailer’s site is designed, the browser method for finding products could take some time.

Searchers, on the other hand, take a more direct approach. They simply enter a search term into a website’s search box and are presented with results matching their search query. This shopping method generally obtains information much faster than using the browsing approach. But, for searchers, there is often a problem with this way of locating information. Namely, the results to the search can contain irrelevant information that does not match what the searcher is really seeking. For instance, for a customer that is truly interested in a regular toaster, the results page of a “toaster” keyword search may include not only regular toasters, but also toaster ovens or even Toaster Strudel (food item). It is because of such ineffective search results that many customers find being a searcher is actually more time consuming than browsing. This may especially be the case when customers use their mobile phones to find products as the small screen space often restricts shopping to using the search method.

Now it appears search results at online retailers are getting better. As discussed in this Internet Retailer story, several companies are promoting search software that is much better in pointing customers to the products they are really interested in seeing. As noted in the story, the new technology often presents searchers with what the software predicts to be the best match to a search query. Thus, in some cases, instead of seeing multiple pages of products, the searcher may only be shown a single product that the search technology estimates is the best match for the query the customer entered. How can search technology know this? Primarily by using “learning” methods that evaluate hundreds-of-thousands, if not millions, of search queries to see how customers behave.

The story presents the site search experience of several retailers, including one retailer who believes it has helped increase sales by more than 20%. The story also discusses the companies offering advanced search technology.