Skip to main content

Making Online Shopping More Fluid and Organic

Many times, shoppers go into stores without knowing what they need or want. Modern consumerism thrives off of this idea. Americans rarely need everything they have.

One of the biggest challenges that online retail stores face is replicating the aspect of shopping that appeals to the hobbyist consumer – the consumer that goes to the mall, alone or with company, to see what’s on the floor, without having a predetermined goal or product that they are seeking out. In a physical mall experience, shops are organically able to draw crowds to their store by visuals in their display boxes, sales, and reputation. The consumer is able to see a large variety of products with different uses before choosing specific products to seek out.

It is difficult to replicate this in an online shop or marketplace because online shopping is search query based. There are very low chances that a user goes to, searches for available leather band watches, and ends up buying snow boots instead.

However Etsy has found a way to overcome this problem by modeling their products as a network. First using the Hub and Authority algorithm to identify the products that will have the most success and then using an algorithm based on the law of entropy to disperse the query results across different categories of products in an appropriate way. It utilizes the assumption that all categories have products that may have attributes in common and is able to predict with category the query pertains most closely to, while also marketing other categories to the user. This limits the possibility of the user having tunnel vision, and allow the user to “window shop” when they are not quite sure what they are specifically looking for.

Etsy uses the network model to not only show related products, but to also predict which product will be more successful, by linking customer reviews, vendor feedback and quality of page, and number of clicks on certain items. It utilizes a ranking algorithm similar but not identical to the PageRank algorithm to which products will be the most successful.

While consumers used to strictly rely on review hosting websites to make their decisions on products they should buy online, they can now rely on the online marketplace itself to bring them relevant and successful products. Furthermore, online shopping has gotten exponentially more fluid and natural over the past decade, thanks to search and web algorithms that allow more connectivity between products, categories, and vendors.  These algorithms are making online shopping more fluid and enjoyable and in the most extreme cases even causing some people to even go on “shopping sprees” online when bored.



How Etsy Uses Thermodynamics to Help You Search for “Geeky”



Leave a Reply

Blogging Calendar

October 2016