How did you begin working together?
Cachon: We started working together when Marcelo was in his second year. He came to me because he was interested in doing empirical work in the field of operations management, and I was also interested in that. I had a few ideas for empirical projects, and we just went from there.

Olivares: Gerard is famous for this kind of work. He has one of the seminal papers in which he uses game theory to analyze problems in supply chain management. I was interested in doing something at the intersection of economics and supply chain management, but more so putting some data into the analysis — basically, bringing some of the tools from econometrics to analyze problems in supply chain management and operations.

Tell us about your collaborative research.
Cachon: There are two papers coming out of Marcelo's dissertation that I was primarily involved with. Both are associated with the auto industry. The first one is looking at inventory holdings of the major automotive manufacturers in the United States over the last 10-15 years. What we've noticed is that Toyota tends to carry much less inventory than all of the other manufacturers. We were interested in trying to discover what was special about Toyota that allowed it to carry much less inventory. That was one project that was at the company level.

Then we decided to look at the micro level of actual inventory at dealerships. It turns out that GM, through their website, allows customers to track inventory for each dealership. Marcelo designed a web crawler to pull data from several hundred dealerships throughout the country. So, over a six-month period, he would query these dealerships to see what cars they had available.

Using this data, we could identify exactly where vehicles were located in the country in the GM dealership network. Then, we were interested in studying how local competition influences how much inventory these dealers carry today and whether local competition makes them carry less or more inventory.

What are the implications of your research?
Olivares: I think there are several implications. The first is academic — we have done research that merges operations management theory with practice. This might be obvious, but if you look at literature in operations management, most of it has been theoretical. There haven't been that many studies that have merged theory with data.

Cachon: One of the reasons why Toyota has lower inventory is they're more flexible in their manufacturing and they carry fewer product lines. This was the first empirical validation that product variety will cause you to carry more product inventory.

The second finding showed that local competition increases inventory. Dealers who are facing more competition have higher inventory. Because GM has more dealerships, their dealerships are going to face more competition. As a result, GM is going to carry more inventory than Toyota. There have been many discussions in the auto industry about whether or not GM has too many dealers. Our findings do not answer that question, but they do add to the question. Part of the differences between GM and Toyota can be explained in the difference of their dealership structure.

Olivares: Both are suggested by theoretical models, but there wasn't much empirical evidence showing that these factors indeed matter in practice. I think that's an important contribution of the work.

What have you learned from working together?
Olivares: The most important thing I have learned from Gerard is to be open-minded when doing research. You shouldn't be too biased about an idea. You should be open to finding the true causes driving a particular phenomenon, instead of trying to push a particular view. You should be open-minded and open to reading different types of literature and try to dig as much as possible into the problem until you get to the root causes.

The other is being extremely rigorous in the work you do — not leaving any single detail hanging out there. If there's anything in your work that you're not sure about, you should dig further to nail down every single possible issue remaining.

Cachon: For me, until I started with Marcelo, I wasn't doing any empirical research. Working with Marcelo motivated me to learn more about empirical methodology and become more interested in it. It definitely made me more data-driven in my own research.

How has being at Wharton benefited your collaboration?
Cachon: Marcelo needed to do a lot of Web data collection, and he used the computing resources extensively in order to be able to collect all of these data. That was a tremendous asset. Other faculty who are also interested in the auto industry provided us with a lot of ideas and suggestions and discussions.

Olivares: One thing I like about Wharton is it provides you with a flexible PhD program. There's a sense of openness, being open to different ideas. You can take courses in marketing, accounting, finance, or economics. All of the functional areas are interconnected, and it's positively viewed if you have a broader view when working on your PhD.

Cachon: In my department, there are many faculty members working on broad range of ideas and issues from very many different methodologies — deep theory people to very data-driven people, people who want to work with companies directly, and others who work with data collected by government agencies or industry consortia. When students come in to our department, we let them come in and work with anyone they want. So there's a lot of flexibility. Marcelo came in thinking he wanted to study information systems, but then gravitated towards studying operations management with an empirical bent.

In addition, some students work with faculty outside of our department. A lot of our students work with marketing faculty or management faculty. Given that Wharton has more than 200 faculty, it would be almost impossible for a student to find a business topic that isn't related to the work of some faculty member at the School.

Will you continue to work together in the future?
Cachon: Marcelo's going to be a faculty member at Columbia, and he needs to make a name for himself. I'm aware that he needs to go out there and get tenure. He'll be working with other people and working on his own projects. But as he's doing that, I'm hoping that I'll be able to carve out some time to work with him on additional projects.

What advice do you have for doctoral program applicants?
Olivares: You have to be open-minded. You have to know that once you get into a PhD program, you will be exposed to a lot of new ideas, a lot of new disciplines, which may change your interests and may give you new ideas. Be prepared that your initial ideas for coursework and research are going to change during the program.

Cachon: Students need to be extremely self-motivated, be able to work independently and relish the challenge of working with a lot of intelligent people — a lot of intelligent faculty, a lot of intelligent students. And students have to come to Wharton with humility. Although they're probably really smart, they're going to be around a lot of other really smart people. They need to thrive on that.