Wayfair Tech Blog

A Senior Economist at Wayfair Provides Three Tips for PhD Students to Land a Highly Selective Internship at the Company

Senior economist Colin Gray plays a leading role in hiring economists from universities across the country for both internships and full-time roles.
Sarah Hill

According to Colin Gray, a growing number of economists are being drawn to the prospect of carrying out interesting work at technology companies like Wayfair.

“As an economist at Wayfair, I spend a lot of time conducting causal analysis,” Gray says. “While this work involves prediction and machine learning, there’s an important nuance. I'm often less focused on prediction for its own sake, and more focused on understanding the causal impact of pulling various business levers.”

Gray says that economists at Wayfair are tackling problems in diverse areas, from causal inference to time series forecasting and price modeling.

Gray was in college during the Great Recession of 2008. He decided to pursue a career in economics to develop a deeper understanding of the forces that impacted so many facets of our lives. Gray was also drawn to the multidisciplinarity of the field. Economists often have to excel in a number of disciplines from statistics to research and communications – think of the Chairman of the Fed communicating the impact of interest rate changes to the media and lawmakers.

At Wayfair, Gray plays a leading role in hiring economists from universities across the country for both internships and full-time roles. Wayfair is highly selective when it comes to hiring interns in economics every year.

In this article, Gray provides three tips for students pursuing PhD degrees in economics to excel during the interview process, and land an internship at Wayfair.

1. Demonstrate an ability to communicate with business stakeholders.

“When you’re trained as an academic, you are taught to communicate with other academics,” Gray says. “However, when you are working at a technology company like Wayfair, being able to communicate with smart business stakeholders who might not have a deep background in statistics is critical.”

During interviews, Gray looks for candidates to demonstrate the ability to take a complex model, and translate it for business stakeholders. After that, the candidate should be able to show that they are excited about being able to understand stakeholders’ concerns, and translate these back into their models.

“You might encounter a situation where a stakeholder wants to get insights into something that’s very statistically noisy. For example, we are often tasked with trying to determine which group has the biggest effect on a particular business metric.”

Gray says that economists often resort to a technique called shrinkage, where they build models that allow an estimate to differ across groups in the data, but only when there is sufficient evidence that a specific group is indeed different. Most statisticians would instinctively know that this is a distinct technique than coming up with a different estimate for each group.

“What I am looking for during the interview is for a candidate to demonstrate an understanding that this distinction is not an obvious one for business stakeholders. Candidates should be able to translate that insight in a simple and understandable way,” says Gray.

2. Have different options for every scenario

Gray says that one of the biggest differences between working in industry and academia is that at a technology firm, you have to carefully determine which modeling improvements will make a meaningful impact on the business.

“When we are building a model in academia, we don't stop until we achieve almost-perfection,” says Gray. “However, working at most technology firms requires a complete change in mindset. Here you want to hire candidates who know which shortcuts matter, and where making incremental changes can have a meaningful impact.”

Gray says that during interviews, he is looking for candidates to come in with different options for the same problem.

“There’s never one right answer. It’s always good to have multiple options with different strengths or weaknesses, since a ‘fast’ technique that's 90% accurate may be a better use of time for the business than a technically pristine method that's 95% accurate,” he says.

3. Be curious and open-minded about alternative methods

According to Gray, after being trained in academia, economists can get into a mindset where they feel that they have a monopoly on the best method to solve a particular problem.

“What we are looking for at Wayfair is for candidates to be curious and open minded,” he says. “It’s really important to demonstrate an ability to broaden your horizons, and pick the best tool for the problem regardless of how you were trained.”

Gray is looking for candidates to display curiosity and some understanding of techniques outside of mainstream econometrics -- such as Bayesian inference, multi-armed bandits, and advanced machine learning. They have to be able to demonstrate that they understand the strengths and weaknesses of each tool, and apply them based on whether they are the best available approach to a problem.

“We are inclined to hire candidates who like to go for simplicity first,” says Gray. “If you add fancy bells and whistles, you should be able to demonstrate why that addition helps you drive towards unraveling complexity, and arriving at a better answer.”

Ultimately, Gray says that the sheer diversity of the problems at Wayfair makes the company an exciting place for economists.

“At a company at Wayfair, you are given the opportunity to solve bizarre problems that you’ve never encountered before. In my opinion, an internship is a great way to get a sense of these interesting problems we face, and become familiar with the day-to-day life of a tech economist.”