Myth #2: Search is catalog-driven only.
Search usually starts with a retailer’s product catalog. A product catalog consists of the items a retailer is selling, as well as all the information associated with each product. Catalogs contain details for each product such as price, description, color, size, nutrition, weight, volume, and other facets. Unfortunately, most search engines don’t ingest all the relevant information associated with each product, making it more difficult to return accurate search results.
While a retailer’s product catalog might be the foundation of search, it doesn’t end there.
Good search depends on:
a.) using historical data to build and train smart search models and
b.) creating a feedback loop that monitors shopper behavior and uses it to improve search results
Some examples of behavioral data that inform effective search engines are:
query terms - the specific search terms used to look for a product
clickstream - the sequence of links a visitor clicks on any given site, including
add to cart - the selection of products a shopper added to an electronic cart
product purchase - selected goods that are ultimately bought
Armed with a dynamic product catalog and shopper behavioral data, we can realize significant improvements to search relevance.