Public school to PhD
Overview
Peter Coles, Airbnb's Head Economist for Policy and Director of Data Science, shares his career journey from public school in Milwaukee through Princeton, Stanford PhD in economics, Harvard Business School professorship, eBay's Data Labs, to leading data science and economics teams at Airbnb. The article highlights how his academic background in market design, game theory, and matching mechanisms directly informed his work on marketplace policy, pandemic response analytics, and academic research collaborations at Airbnb.
What You'll Learn
1
How academic economics expertise in market design translates to practical data science leadership at a tech marketplace
2
Why matching theory and market design principles are valuable for platform companies like Airbnb and eBay
3
How Airbnb structures its data science and economics teams to address policy, strategy, and academic research
4
How to build academic-industry research collaborations while respecting privacy and legal limitations
Prerequisites & Requirements
- Basic understanding of marketplace economics and platform business models(optional)
Key Questions Answered
What is the career path of Airbnb's Head Economist and Director of Data Science?
Peter Coles studied math at Princeton, earned a PhD in economics at Stanford focusing on game theory, taught at Harvard Business School where he co-taught with Nobel Prize winner Al Roth on market design, led the economics and data science team at eBay's Data Labs, then joined Airbnb in 2015 to build their economics team for policy analysis.
What is the Central Strategy and Insights (CSI) team at Airbnb?
CSI is a team co-founded by Peter Coles and Jackson Wang at Airbnb to answer big cross-organizational questions that no specific data science team could address alone. The team acts as forensic investigators, piecing together stories from evidence across the company. They addressed pandemic-driven changes in guest demand and supply needs, led business reviews, and generated analyses for shareholders ahead of Airbnb's IPO.
How does Airbnb collaborate with academic researchers on data science?
Airbnb developed a program to collaborate with external academic researchers while respecting privacy and legal limitations. They have published papers with professors from MIT, Berkeley, Stanford, UCLA, and NYU. They also launched a monthly seminar inviting academic collaborators to discuss research with Airbnb data scientists and technologists, fostering cross-pollination of ideas.
How did Airbnb's data science team respond to the COVID-19 pandemic?
During the pandemic, Airbnb's CSI team addressed major shifts in where guests were looking to stay and analyzed the supply needed to accommodate changing demand patterns. They used data-driven analysis to understand these adjustments and inform company strategy during a period of unprecedented disruption to the travel industry.
What is matching theory and how does it apply to marketplace platforms?
Matching theory involves mechanisms to pair users from two groups, often when price alone cannot clear the market. Peter Coles researched participant strategy, signaling in markets, and even improved the market for PhD economists. This academic foundation in market design directly applies to platforms like Airbnb and eBay, where matching guests with hosts or buyers with sellers is the core challenge.
What was the 'What's it Worth' project at eBay?
What's it Worth was a project Peter Coles worked on at eBay with colleague Dean Chen to develop a methodology for determining the fair market value of items sold on the platform. It combined hands-on practical work with economic modeling, representing the kind of applied marketplace research Coles had been seeking after transitioning from academia.
Key Statistics & Figures
Duration at Airbnb
Eight and a half years
Peter Coles's tenure at Airbnb at time of writing, spanning three distinct phases of work
Key Actionable Insights
1Transitioning from academia to industry can be highly effective when your research specialty aligns with the company's core business challenges. Peter's expertise in market design and matching theory was directly applicable to marketplace platforms like eBay and Airbnb, making the transition natural and impactful.This is particularly relevant for economists, data scientists, and researchers considering industry roles at platform or marketplace companies.
2Building cross-organizational data science teams (like Airbnb's CSI) that have visibility across the whole company can address strategic questions that siloed teams cannot. These teams serve as forensic investigators, synthesizing evidence from multiple departments to inform executive decisions.This organizational pattern is valuable during periods of rapid change, such as pandemic response or IPO preparation, when holistic business understanding is critical.
3Establishing structured academic-industry research collaborations requires solving privacy and legal challenges first, but once in place, they create significant value through published research, monthly seminars, and cross-pollination of ideas between practitioners and academics.Companies with large proprietary datasets can attract top academic talent by creating formal collaboration frameworks that respect data governance while enabling meaningful research.
4Giving data science teams a mandate to work on ambitious, long-term projects — even those that might take a year or two to prove out or might fail — fosters innovation and attracts top talent who want space for creative problem-solving alongside applied work.This approach to team culture helps retain PhDs and senior scientists who value intellectual freedom alongside practical impact.
Common Pitfalls
1
Attempting to build a marketplace without understanding the fundamental dynamics of supply and demand matching. Peter's childhood rock stand failed because there was no real demand — neighborhood kids could find their own rocks for free, illustrating that marketplace value requires differentiated supply.
This anecdote underscores that marketplace founders and data scientists need to deeply understand what creates value for both sides of a platform before scaling.
2
Staying too long in a career phase that doesn't fit your strengths or interests. Peter recognized that his short attention span wasn't suited for the long cycles of academic research papers and peer reviews, prompting a well-timed transition to industry where the pace better matched his working style.
Self-awareness about career fit is important for data scientists deciding between academic and industry paths.
Related Concepts
Market Design
Matching Theory
Game Theory
Data Science Leadership
Marketplace Economics
Platform Quality
Economic Impact Analysis
Academic-industry Collaboration
Two-sided Marketplaces
Sharing Economy