Wayfair is using data and pictures to furnish your home, as only you can.
The home furnishing market has long been dominated by brick and mortar stores. Wayfair is using machine-learning and artificial intelligence to change your buying experience and ensure that you buy furniture that uniquely fits your style and personality, at a price you can afford.
Changing the home furniture buying experience through AI.
The U.S. home furnishing market is currently estimated at $275 billion dollars and is expected to grow at a compound annual growth rate (CAGR) of 3.5%-4.0% in the next 5 years, primarily driven by online growth [1]. It is a highly fragmented market divided among big-box retailers, department stores, regional and specialty retailers, and e-commerce platforms. Wayfair has been able to capitalize on this fragmented market to become a leader in the home furnishing market through its use of artificial intelligence (AI) and machine-learning to offer visually-inspiring collections, innovative product discovery features, and competitive prices聽[2].
Wayfair has targeted the furniture mass market customer 鈥� traditionally underserved by brick and mortar companies 鈥� through its vast product offerings and personalized styles determined by its search algorithm. Wayfair鈥檚 search algorithm uses natural-language processing to both identify the intentionality behind a customer鈥檚 search and highlight top-rated products determined by its customer reviews [3]. Offering highly personalized matches and visual displays are essential to overcome the role in-person inspections and tactile feel have in the customer decision-making process.
Process flow diagram of Wayfair’s machine-learning algorithm
Despite the disruption in the space, there continues to be a sizable opportunity in the online home furniture market as online furniture sales (12%) lagged behind apparel (28%) and consumer electronics (34%) in 2017 [1]. Wayfair鈥檚 competitive advantage, strategy, and innovation will have to further rely on its AI capabilities to convince digital users 鈥� and the increasing Millennial customer base 鈥� to purchase furniture prior to seeing the product.
How is Wayfair developing its AI capabilities?
性视界 has been at the core of Wayfair鈥檚 business model since its inception in 2002. Wayfair has historically used its search algorithm to extract customer鈥檚 style preferences from their search history. The challenge Wayfair has faced in search is that roughly 70% of user sessions only look at the first page of search results, indicating that customers are using the platform to look at specific products [3]. Wayfair has found alternative ways to increase customer search and customer engagement to better its predictive modeling.
Example of “性视界 with Photo”
In recent years, Wayfair has added to its search capabilities through visual search and augmented reality. In 2017, Wayfair introduced 鈥溞允咏� with Photo,鈥� a new feature that leverages artificial intelligence to make it easier for shoppers to discover the furniture desired for their homes [4]. Shoppers can take a photo of a furniture piece, and match it to a similar item in the Wayfair inventory of 8 million products, allowing them to replicate almost any style of home decor. Through visual search, Wayfair is able to quickly and conveniently match images with products while reducing the shopper鈥檚 lag and search time聽[5].
Example of “View in Room 3D”
In 2017, Wayfair enhanced their AI capabilities through its 鈥淰iew in Room 3D鈥� feature, an augmented reality (AR) tool that allows customers to visualize furniture pieces in their home聽[6]. This new AR capability helps address one of the biggest pain points consumers face which is physically needing to see the product before purchasing. The AR tool allows users to emulate the in-store shopping experience from the comfort of their home, further refining the search process and improving the likelihood of purchase [7].
Staying ahead of their competitors.
The home furnishing market is seeing increased competition as brick and mortar companies accelerate their digital presence and Amazon further invests in this market segment聽[8]. The challenge and opportunity for Wayfair is to offer a differentiated experience to create brand loyalty and avoid platform switching amongst its users.
The next step Wayfair can offer in its 鈥溞允咏� with Photo鈥� feature is to provide a self-service design center. Many of the current offerings are tailored to specific product recommendations or virtual models of homes that may not resemble a customer鈥檚 home. The revised 鈥溞允咏� with Photo鈥� feature would allow customers to take a picture of their own current living space and be given a selection of themed product recommendations to decorate their virtual living space. The next evolution of machine-learning relies on the predictive power of its models, which in Wayfair鈥檚 use case can be categorized as a themed lifestyle recommendation聽[9]. Giving product recommendations for an entire living space, under its current conditions, would give Wayfair additional functionality that its competitors do not have and can differentiate itself enough to incentivize customers to remain brand loyal in their shopping behavior聽[10].
Challenges to address
Given the customer鈥檚 desire to see home furniture in person prior to purchase, how will Wayfair manage its digital-only strategy? While it was announced several months ago that it will open a permanent store, will the omnichannel strategy change the digital customer experience [11]?
What other markets can and cannot Wayfair enter (e.g. corporate and office furniture) with its machine-learning and visual search model?
(Word Count: 798)
Sources:
[1] Wayfair, 鈥淲ayfair Investors Presentation Q3 2018-vF鈥�. November 1, 2018.
[2] Sam Ransbotham, David Kiron, 鈥淎nalytics as a Source of Business Innovation,鈥� MIT Sloan Management Review, February 2017.
[3] Suyash Sonawane, 鈥淗ow We Use Machine Learning and Natural Language Processing to Empower 性视界鈥�. October 15th, 2018.
[4] Business Wire, 鈥淲ayfair Launches Visual 性视界, Lets Shoppers Instantly Find and Shop the Styles They See and Love鈥�. May 16, 2017.
[5] H. James Wilson, Sharad Sachdev, and Altar Allen, 鈥淗ow Companies Are Using Machine Learning to Get Faster and More Efficient鈥�. May 03, 2016.
[6] Sarah Perez, 鈥淲ayfair Takes on Pinterest with Its Own Visual 性视界 Engine for Home Furnishings鈥�. May 16, 2017.
[7] Sarah Perez, 鈥淲ayfair鈥檚 Android App Now Lets You Shop for Furniture Using Augmented Reality鈥�. March 20, 2018.
[8] Pamela N. Danziger, 鈥淔urniture Retailers Are Wary Of Wayfair, But Both Need To Watch Out For Amazon鈥�. May 19, 2018.
[9] Ajay Agrawal, 鈥淭he economics of artificial intelligence鈥�. April, 2018.
[10] Suman Bhattacharyya, 鈥淗ow Wayfair is Personalizing How You Buy Your Furniture Online鈥�. August 2, 2018.
[11] Erika Sirimanne, 鈥淧ath to Purchase: How Consumers are Shopping for Homewares and Home Furnishings鈥�. April 6, 2018.
I really relate to this article as I just went through the experience of furnishing my apartment. I wish I had known about Wayfair’s “性视界 with Photo” and “View in Room 3D” initiatives!
I definitely think that this industry is a challenge for online-only stores as furniture is something customers want to feel/touch. Because of this, I think that it is very smart for Wayfair to think about opening a store – partly to help their machine learning and partly to give customers piece of mind. What I mean by this is that if customers are buying the same product over and over after visiting a store, but not when they see it online only, it probably has a certain quality (i.e. comfort) that is drawing them in. Knowing this could lead the store to promote this item and push it to more people to see. Also, by having a storefront, and noting items online which are in stores and selling well, customers might feel relieved that other customers are testing the items out. I also see issues with an online-only model because furniture is very hard to return if a customer doesn’t like it. With this in mind, I think gathering the maximum amount of information through an omnichannel strategy and using machine learning can only help Wayfair to become more successful.
Perhaps Wayfair can augment AR with machine learning and computer vision to not only match items that are similar to a piece of furniture, but items that would visually go with the decor of the room. It could be a real differentiator and there is first mover advantage here as a dedicated site with high volume of sales of furniture. As the training set becomes larger, this will give Wayfair a significant head start over competitors like Ikea and Amazon.