Wayfair Tech Blog

Wayfair’s New Approach to Aspect Based Sentiment Analysis Helps Customers Easily Find “Long Tail” Products

Over thirty million customers from around the world shop on Wayfair to look for furniture that helps them create their unique sense of home. Wayfair serves a variety of webpages catered to niche customer sub-segments and help customers discover new products – these include pages with advice for people buying “twin futons,” or “nursery chairs with removable cushions.”

Wayfair Tech Blog

Wayfair Intern Shima Dezfulian: Everyone Here Wants to See You Succeed

Like many interns at Wayfair, Shima Dezfulian works on projects that have a high impact on the company’s operations. More specifically, Dezfulian is part of a Wayfair’s finance and algorithms team that leverages machine learning to make predictions about Wayfair’s profitability.

Wayfair Tech Blog

Utilizing Piecewise Linear Approximations – With a Twist – to Make Supply Chain Decisions

The annual conference for the Institute for Operations Research and the Management Sciences (INFORMS) brings together professionals and experts from the industry and academia. At INFORMS 2022, I will be giving a presentation on Global Network Optimization (GNO), which is an ongoing project carried out by Wayfair’s operations research team.

Wayfair Tech Blog

Wayfair paper accepted at RecSys 2022 applies state-of-the-art method to generate product recommendations

Style transfer: How Wayfair leverages customer browse activity in one furniture category to inform recommendations for a category they have never shopped before.

A robot icon with a word bubble saying "Hi, there! I'm Wayfair's virtual assistant. Let me know how I can help!"

Wayfair Tech Blog

Building Wayfair’s First Virtual Assistant: Automating Customer Service by Text Based Intent Prediction

Not everyone wants to speak with someone when they need to set up a return or report an issue with their order. We built an NLP-powered Virtual Assistant to provide our customers a fast and easy option that’s available 24/7 to resolve their service-related issues.

Product detail page for a sofa showing ratings as well as 8 user-contributed images

Wayfair Tech Blog

Predicting Image Quality in Product Reviews

Images from product reviews can inspire and inform our customers. We created a machine learning model that helps us in prioritizing the most helpful and good looking images on the product page.

Flowchart which shows customer activity feeding into models to estimate the value of sending an email or push notification, optimizing the time to execute the touchpoint, and then make a final determination to send or not send

Wayfair Tech Blog

Nightingale: Scalable Daily Sales Email Sending Decision Model

Wayfair and its subsidiary brands operate globally and we send out millions of marketing emails every day to communicate with our customers. To improve customer satisfaction and increase customer engagement, we have developed a new generation of daily sales email models (Nightingale) to make email sending decisions. The Nightingale model is built via a scalable retraining pipeline to serve different business needs across all regions and brands.

Wayfair Tech Blog

An Inside Look at Wayfair’s Data Engine, Scribe

Today, many companies rely heavily on third-party event tracking platforms for collecting event data. Perhaps the most well-known of these is Google Analytics. However, at Wayfair, we chose to go another route, largely eschewing the third-party vendor model to develop our own first-party event tracking platform. It’s called Scribe.

Count of orders place with Wayfair per day, showing weekly, seasonal, and shock fluctuations

Wayfair Tech Blog

Time Informed Calibration

The Time Informed Calibration process extends the capabilities of Wayfair’s customer models by converting raw model output to realistic probabilities and incorporating information on expected trends.