Keywords based eCommerce search employed by most retailers today is inherently flawed. In most cases, it has become a dumb word-matching engine that tries to come up with closest match given a set of words. It does so without understanding the context of who is using it, what is it being used on and what are other customers saying about the products that are being searched on.
So, if you go to an ecommerce store and try searching with keywords ‘pregancy coats’ you will be shown anything that is either related to pregnancy or coats. In our experience, we have seen top search results that range from a nail polish that gives you an even coat of polish to a herbal tea that gives your throat a nice coat to make your sore throat better to a book by an author whose last name is coats.
If your customer who was looking for coats to wear during pregnancy saw these search results, she would assume that you do not carry any relevant products and move on to the competitor’s website.
Have you noticed that your search engine returns results for Zombi when your customers search for Zumba.
You search for shoes with good arch support and the search engine returns the shoes which is mostly negatively reviewed for arch support. It might not be relevant but hey, the keywords match
Searching for Blues Guitar returns the guitar which is blue in color on your music instruments website. Your customers are thinking - "Are you kidding me?"