TL;DR:
- Wayfair employs ML technology to enhance the online search experience and ensure customers find the exact products they need.
- The specially designed platform, developed in collaboration with Snorkel AI, enables accurate and efficient programmatic labeling.
- Wayfair’s ML models consider various characteristics of products, resulting in higher conversion and customer satisfaction rates.
- The partnership between Wayfair and Snorkel fills the gap in programmatic labeling for images, complementing Snorkel’s existing text-centric platform.
- Wayfair streamlines catalog management by empowering subject matter experts to quickly create new product categories and automate the labeling process.
- The programmatic approach allows Wayfair to respond swiftly to emerging home furnishing trends and make catalog updates efficiently.
- Wayfair envisions expanding the use of programmatic labeling for other applications, driving efficiencies, and leveraging data for innovation.
Main AI News:
Wayfair, the prominent home-furnishing specialist, has leveraged the power of Machine Learning (ML) technology to revolutionize its online shopping experience. By employing a purpose-built platform, developed in collaboration with Snorkel AI, Wayfair aims to elevate the quality of its search functionality and ensure that customers discover precisely what they are seeking. Tulia Plumettaz, Director of Machine Learning at the e-commerce giant, sheds light on this innovative approach.
The platform enables Wayfair’s customers to effortlessly find products by using their own terms. Plumettaz provides a specific example: imagine searching for a “cerulean sofa.” By simply typing those words, customers can expect the exact results they desire. This ability to describe products in their own words significantly boosts conversion rates and customer satisfaction.
Addressing the underlying business challenge, Wayfair frequently delivers bulky items like sofas to its customers. Plumettaz explains that the company sought to bridge the gap between the information received from suppliers and the customer’s perception of the product. By employing ML technology, Wayfair endeavors to provide customers with an accurate representation of the product they will receive. Plumettaz elucidates, “We pondered whether we could build models based on the characteristics of a sofa. Does it feature a chevron pattern? Is it yellow? What is the texture? Are the arms rounded? These are the aesthetic details customers use when searching in their own terms.”
Building these models necessitates an extensive amount of training data. However, Wayfair strives to minimize the burden on suppliers while maintaining efficiency. The challenge lies in constructing models swiftly, cost-effectively, and with accuracy, which is precisely where Snorkel’s solution comes into play. Plumettaz emphasizes that Snorkel facilitates fast labeling operations, or programmatic labeling, enabling accurate model training with improved speed and cost efficiency.
Wayfair’s collaboration with Snorkel is classified as a “design partnership.” Initially, Snorkel’s key product, Snorkel Flow, primarily focused on text-related applications. However, Wayfair required assistance with the programmatic labeling of images, which led to an amalgamation of efforts between the two companies. Plumettaz notes that Wayfair sells images as products, making it vital for Snorkel to venture into this new market encompassing programmatic labeling for computer vision. The partnership enables Wayfair to influence the product roadmap, resulting in a unique offering tailored to their needs.
Plumettaz describes the relationship with Snorkel not as a typical vendor-client arrangement but as a collaborative journey where mutual interests align. This unorthodox partnership showcases the significance attributed to each other’s success. The collaboration spans over a year, during which Wayfair sought to develop an in-house solution. Recognizing that ML technology development was not their core business activity, Wayfair sought an external partner. Serendipitously, Wayfair’s CTO encountered Snorkel at a conference, sparking discussions that culminated in this fruitful partnership.
The resultant technology stemming from this collaboration is deployed to enhance and refine Wayfair’s catalog. Leveraging Snorkel’s platform, Wayfair adopts a data-centric approach rooted in ML best practices and cutting-edge foundational models. Plumettaz explains that subject matter experts within Wayfair can easily create new product categories and ensure that the appropriate images are presented to consumers.
With just a few clicks, these experts can generate a preliminary depiction of a new category, enabling Wayfair to make strategic decisions quickly. If successful, the process can be automated and integrated into production. This empowers category managers and expands the user base beyond data scientists to any personnel involved in image labeling.
The programmatic approach, now an integral part of Wayfair’s operations, simplifies catalog modifications when emerging home furnishing trends emerge. Plumettaz highlights the significant impact of this advancement: “Before Snorkel, we relied on humans to painstakingly label 40 million products. Now, with programmatic labeling, we can swiftly relabel our catalog, aligning it with the new styles customers are discussing. This brings true value to our business.”
Furthermore, Plumettaz envisions applying this technology to other use cases. From providing more detailed information to marketing efforts to reducing product description requirements for internal teams, Wayfair intends to exploit its partnership with Snorkel to drive efficiency and foster innovation. Plumettaz underscores the pivotal role of data in this endeavor, noting that programmatic labeling will become an indispensable tool for achieving meaningful business transformation at scale.
While acknowledging the significance of technology in implementing machine learning processes, Plumettaz stresses that finding a strong use case is equally vital for business leaders pursuing a data-led strategy. The technology must drive business value and serve as a differentiator, rather than being implemented solely because it is perceived as trendy. Plumettaz advises organizations to anchor their conversations on how technology can genuinely enhance their business.
Wayfair’s strategic collaboration with Snorkel has propelled the company to new heights in the realm of machine learning implementation. By harnessing the potential of programmatic labeling, Wayfair not only ensures an exceptional customer experience but also optimizes its catalog management processes. As the partnership continues to evolve, Wayfair anticipates leveraging data-driven approaches in diverse areas of its business, fostering efficiency and delivering unparalleled value to its customers.
Conclusion:
Wayfair’s collaboration with Snorkel AI to implement machine learning technology represents a significant advancement in the market. By utilizing programmatic labeling and a data-centric approach, Wayfair streamlines its catalog management, enhances customer experience, and fosters operational efficiencies. This strategic partnership not only allows Wayfair to meet the evolving demands of the home furnishing industry but also demonstrates the power of ML technology in driving business value. As data continues to play a pivotal role, programmatic labeling is poised to become an essential tool for businesses seeking to differentiate themselves and achieve success at scale.
