Etsy explained how it uses recommendations to help shoppers narrow down their search for special items on its platform that features over 100 million unique listings.
The recommendations are ubiquitous across Etsy, according to a new post on its Code as Craft blog for developers (written by Etsy engineers), which said the recommendations are tailored for different stages of a user’s shopping mission.
“We call each recommendation set a module, and there are hundreds of them both on the web and on mobile apps. These help users find trending items, pick up shopping from where they left off, or discover new content and interests based on their prior activity,” according to the post.
Here’s how Etsy figures out which listings to recommend in any given situation (in a nutshell):
“Modules in an enterprise-scale recommendation system usually work in two phases: candidate set selection and candidate set ranking.
“In the candidate set selection phase, the objective is to retrieve a small set of relevant items out of the entire inventory, as quickly as possible.
“The second phase then ranks the items in the candidate set using a more sophisticated machine learning model, typically with an emphasis on the user’s current shopping mission, and decides on the best few items to offer as recommendations.”
The post goes on to describe the latter, explaining that it uses contextual attributes, such as the user’s recent purchases and most clicked categories, as well as item attributes such as an item’s title and taxonomy, and then optimizes the ranking process against a specific user engagement metric, such as clickthrough rate or conversion rate.
The post includes a screenshot showing two recommendation modules on a listing page, explaining that the “more from this shop” module recommends similar items from the shop the user is currently looking at, while the “you may also like” module finds relevant items from across Etsy shops.
The engineers revealed they had begun running experiments live on the site in the second quarter of last year to test its approach to improving its recommendations.
“Moving forward, we’ll continue iterating on this ranker to improve our target metrics, making the ranker more contextual and testing other novel model architectures,” the Etsy engineers added.
I am so sick to death of Etsy trying to read the shopper’s mind! The only thing they need to focus on is developing the easiest, clearest, visual page with impeccable search result options.