A Case Study of Two Leading Retailers
Jess Erdman, September 2021
All eCommerce retailers know it: modern day fashion discovery has a problem. In fact, it’s such a huge problem that eCommerce retailers are missing out on revenue and customer loyalty.
And, it’s not a new problem. Fashion discovery is difficult for customers, who want to find products that fit their criteria.
And, fashion discovery is just as difficult for retailers, who are left to spend hours updating unreliable product tag information from suppliers.
The way that retailers and customers are experiencing fashion discovery and search is in desperate need of an update.
Through hundreds of interviews, we discovered the most important criteria to shoppers: fit and style. When a customer begins to narrow down their search using more and more keywords, those keywords are related to the fit and style of a product.
Customers that prefer a loose-waist rarely will want to see search results of a tightly-fitted waist.
So, why is it that current fashion discovery tools allow for inaccurate search results? When a customer types in “A-line dress,” she might receive dresses that aren’t cinched in the waist.
The answer: text-based search and fashion taxonomy.
Fashion taxonomy is defined as the categorization of fashion products. If the term sounds complicated to you, imagine how complicated it would be for a customer to know which fashion terms to type into a retailer’s website.
The second piece of the equation, text-based search, relies on customers to understand fashion taxonomy and to guess the terminology that retailers use (for example, an A-line dress might be called a skater dress for some retailers).
In short, customers know their preferences when it comes to fit and style, even if the customer is in “fashion discovery” mode.
And, text-based search allows for too many inaccuracies, and too much confusion around fashion jargon.
We talked to two leading fashion retailers in Korea, Kolon Mall and Lately, who were both plagued with the same problem: inaccurate, burdensome search for customers as well as the burdensome task of product tagging each and every new product that arrives.
With thousands of hours of productivity lost to the task of product tagging, Kolon Mall and Lately wanted a new way to engage with search, from both a customer and retailer perspective.
Enter the YesPlz Style Filter.
The YesPlz Style Filter is a fashion discovery tool that’s powered by artificial intelligence, and eliminates the need for confusing text-search. Instead, customers simply select their preferred style attributes on a virtual mannequin, and receive instant results in real-time that are accurate and fashion-forward.
On the retailer end, the YesPlz Style Filter algorithm learns retailers’ products, automatically tagging the most important fashion attributes along the way.
When Kolon Mall and Lately implemented the YesPlz Style Filter to reimagine the fashion discovery process, both retailers saw an increase in customer satisfaction and more accurate search results.
You can read the full case study here: How Fashion AI Transformed Two Online Retailers
Visual search is disrupting fashion eCommerce, creating new opportunities for retailers whose search is lacking. Powered by fashion AI, visual search has been embraced by major retailers like Shein--but what makes it so revolutionary?
Service planning is hard when you’re trying to integrate technology with service offerings. In this guide for fashion eCommerce, we go over step-by-step, how to plan a new service using fashion AI. You can also discover our fashion search templates to help you along the way.
Fashion search is everywhere, but there are plenty of examples of poorly designed, difficult-to-use search. In this article, we'll break down all of the components that make great fashion search: product filters, image tagging, and intuitive search--and provide you with a checklist to evaluate your own search.
Your request has been submitted successfully.
Our representative will reach out to you within 24 hours.
Contact us at firstname.lastname@example.org for any further questions.