Wayfair is an e-commerce company specializing in home goods. A quick glance at their sleek website shows that the shopping experience here is definitely different from that of Amazon or Walmart. Pinterest-esque panels showcase attractive furniture arrangements, allowing you to pick items out visually instead of just searching by a specific item.
On Episode 36 of AI at Work, we’re joined by Dan Wulin, head of data science and machine learning at Wayfair, to take a look behind the scenes of the elegant storefront and hear about how AI will continue to remodel te online shopping experience.
1. How do you think about some of the differences between Wayfair and Amazon or Walmart or other e-commerce players?
What sets Wayfair apart from other online retail giants like Amazon and Walmart is both what’s sold and how it’s sold. “The nature of what we're selling is fundamentally different from what core products could look like for some of our competitors,” explains Dan. Many different people might be buying the same paper towels, but few will have the exact same dining room table or chandelier in their home. So between product selection and efficiency around shipping larger items, Wayfair is distinct.
From a data science perspective, visuals are key. “A lot of our customers would have a hard time articulating what exactly they're looking to buy when they get to the website,” says Dan. “That can be because they're just not familiar with what furniture style or home decor style actually is, how to put that into words.” Inter-team collaboration at Wayfair focuses on creating a more frictionless user experience.
2. How would you describe the structure of the data science team at Wayfair? What is the interaction like with other parts of the company?
“We’re very much a hybrid organization,” says Dan, whose team consists of around 100 data scientists. Centralized reporting lends uniformity to recruiting and training, while an element of decentralization embeds specific groups within different business functions. Dan explains that the co-location of groups of data scientists with subject matter experts on the business end of things allows fruitful collaboration on projects.
The exchange with engineering folks is likewise fluid. “The guiding principle is that we want to do whatever we need to get the work out in a high quality state as quickly as we can,” says Dan. Sometimes the handoff to the engineering team may be clear, and other times less so, and the data science team stretches to fill the gap or hires someone with the necessary skills.
Over time, as Wayfair grew, the profile of the data science team transitioned from primarily generalists to greater specialization, as the earlier low-hanging fruit type of problems have been addressed and the organization’s needs have evolved. “In some areas, to be able to get the ROI that we're looking for, we need a higher level of specialization. So examples of that could be people that know natural language processing, or our experts in computer vision, and that sort of thing,” Dan explains.
3. Looking ahead 7 to 10 years, what are some things that are coming and how will they change the online shopping experience?
“One problem that we’re working on is understanding style,” says Dan. Wayfair has a function which lets you submit a photo, and algorithms come up with items which look stylistically similar to that.
“The real problem that you want to solve,” he explains, “is understanding complementarity. A customer comes in, and in reality what they have is a room that they're probably trying to furnish and add more to, and they want stuff that goes well. Whether it's with the wallpaper, or the sofa that they already have, or whatever it is.” Ultimately, this boils down to recommendations that are a lot more relevant.
“We're able to do that because of technology around GPUs, and deep learning, and different techniques that let us use things like imagery and text information,” says Dan. “We're able to use that at scale, in real time as people are browsing, to create really differentiated models and experiences.”
4. Advice for interacting with non-technical executives around machine learning and data science?
“If I had to distill it down a few things, I would say the first thing that I always try to do is demystify what data science is,” says Dan. “Because on the one hand, you can say, well it's this complicated math/statistics and you need a certain kind of background to be able to do it, and you need to be able to code, and so on. When in reality, the truth of it is, what a data science algorithm is doing is not that different from what you would do in an Excel spreadsheet if you knew some statistics and had a huge amount of time. It’s doing it in an automated way at scale.”
In addition to making people feel more comfortable by taking away some of the perceived complexity, Dan adds that another important step is working backwards from what the business is looking to achieve. By starting from the end result, Dan says, people are going to understand the approach very naturally.
5. Are there advantages to being an early adopter when it comes to AI?
“I'm a firm believer that data science done right is a very virtuous cycle,” says Dan. Ten years ago, Wayfair was investing heavily in business intelligence and getting the data itself in a very clean and accessible spot. Over time, this created the foundation for rapid development. For this kind of growth to take place in a company, “you can’t go from zero to one overnight,” says Dan, “you have to build up to it.” Translating to the adoption of AI, building a foundation and getting started ahead of the curve may lend advantages further down the road.
Dan shares some parting thoughts that what excites him the most: seeing instances of AI getting folded into how businesses actually operate, where it is able to influence the decisions they’re making or the customer experience. His strategy to staying on top of latest advances? Face to face chats with people in the field.
Tune in to AI at Work on iTunes, Spotify, Google Play Music, Stitcher, or SoundCloud and share with your network! If you have feedback or questions, we'd love to hear from you at email@example.com or tweet at us @talllainc.