Skip to main content

Revolve Spruces Up E-Commerce With Donde’s ‘Trends’ Feature—Here’s Why

The COVID-19 crisis has upended apparel trends left and right as most shoppers adhere to stay-at-home guidelines, work from home and avoid social gatherings. Such a change in consumer habits is a drastic one for retailers to catch up with, especially as consumers themselves might not know exactly what they are looking to shop for.

With apparel retailers having less of a handle on their clientele’s preferences, Donde Search, an artificial intelligence-powered visual search navigation solution, has unveiled a “Trends” feature within its platform to help empower merchants with a discovery solution for shoppers who need inspiration to buy.

E-commerce apparel and lifestyle brand Revolve is the first retailer that has implemented the Trends feature within its business after seeing success with the Visual Search Navigation feature last year. With Visual Search Navigation, Revolve saw conversion uplifts of 11 percent to 16 percent in “key categories,” which gave the retailer the confidence that Trends could do more to address inspirational shoppers.

“For many fashion retailers, a difficulty that is commonly faced is identifying trends based on visual attributes (color, shape, style) and effectively learning from style-based customer behavior,” said Liat Zakay, founder and CEO at Donde Search. “Our new Trends solution is a winning point for both sides. The retailer has a powerful unified tool to connect both inspirational discovery and specific search, and shoppers will be able to be inspired with new and trendy items. We are confident that Trends will increase shopper’s visual discovery engagement and will lead to a shorter time to find that perfect product.”

Related Stories

By implementing the new Trends feature, merchandisers can leverage performance results on every trend (by the attribute level) and see the most popular searches, views and purchases of the trendy attribute when combined with other attributes. For example, they can find out that a balloon-sleeve blouse is more popular when appearing with a solid pattern and a mint color versus floral and black. Adding Trends to Visual Search Navigation also aims to simplify the clutter of the many menus throughout an e-commerce site.

Overall, the Donde technology is designed to enable retailers to react faster to trends, reduce operational overhead, shorten the time to purchase and improve the user experience. The company claims its platform has “significantly” increased its retailer partners’ sales conversions and average order value by up to 20 percent.

The company leverages computer vision and AI to automatically extract visual attributes from product images and turn them into style-based data that improves merchandising, personalization and search across e-commerce platforms.

Visual search caught the interest of shoppers well before the COVID-19 pandemic. December 2018 research from ViSenze, a provider of intelligent image recognition solutions, found that 62 percent of Gen Z and millennial consumers want visual search capabilities that can enable them to discover and identify products using their mobile devices.

In an April survey, ViSenze said 43 percent of consumers are not able to get the right search results when looking for products, while another 39 percent are overwhelmed by too many search results. Up to 18 percent simply do not know how to describe the product they want. These are all indicators of the difficulty converting customers online when they have no idea how to search for the product they want. With apparel stores sidelined over the past few months, retailers must find better ways to deliver these visuals to their consumers.

According to a January study from global insights and strategy firm National Research Group (NRG), the majority of visual search (81 percent) centers on using a smartphone camera to find more information on a product. It can also be used to find links to products and services in photos and videos online (35 percent) and search the internet using an uploaded photo (33 percent).