Airbnb data scientist Erin Coffman (M.A. ’09, Ph.D. ’12) has transformed her company’s culture by making data literacy a priority.
BY TORIE ROBINETTE
As Airbnb’s most senior data scientist, Erin Coffman leads the company’s philanthropic team — a unit focused on housing refugees, victims of natural disasters and those traveling for medical reasons. Her role depends on data, which she calls “the customer voice.”
In 2016, the company was growing rapidly, and Coffman realized a need for more data literacy among employees. So, she helped introduce a concept that would prove transformative: Data University (Data U), an internal training academy at Airbnb through which employees can learn how to collect data and leverage it in their jobs.
We chatted with her about the value of data and how organizations can follow Data U’s lead to drive future success.
How do you work with data in your role?
I help the leadership define metrics for success, forecast targets and goals, build data pipelines, design, monitor and interpret experiments, and tell stories with data.
What inspired Data U?
The original idea was born out of necessity. When I started working at Airbnb in 2011, I was supporting a team of around a dozen managers. The team quickly grew to hundreds while I remained the sole data support. There were also a number of people on the team who were eager to learn new skills and knew our customers in a way others didn’t.
I created a program called “SQooL” to teach them how to write SQL [Structured Query Language], both to help me answer some ad hoc questions that were coming in and to give them a way to do their job more efficiently. The earliest group of students became Airbnb’s first set of product specialists, a role that bridges customer service and product development to ensure responsiveness to what users want and need.
Tell us about the range of topics covered in Data U’s 30-plus courses.
We cover theoretical topics in the 100 series (data-driven decision-making and project proposals), applied topics in the 200 (spreadsheets and data visualization) and more advanced topics in the 300 (R and Python for statistical programming and data visualization). There’s been consistent demand for classes around the world at all of our offices.
You’ve spoken before about creating “citizen data scientists,” meaning people from all over an organization who each know how to use data in their work.
It’s important for any decision-maker in the company to be able to ask the right questions and find the right answers. Data is the voice of our customers. With millions of users, it’s impossible to speak with each one individually, but we can use data to reveal their preferences and discern trends.
How can companies make it a priority for their employees to learn basic data science?
A company needs to invest in making trainings relevant and connected to the company’s issues, tools, products and so on. It can be a big upfront investment, but at Airbnb, we’ve seen that it’s worth it. In a fast-growing company, individual employees might not be able to make adequate time for teaching and practice. It works best when leadership prioritizes learning for their teams.
What does it mean to democratize and scale data?
We think of democratizing data as the intersection of having the right set of tools to analyze and visualize data, offering data education to teach people about the tools and giving people the right levels of access so they can get the answers they need for their jobs while always protecting our users’ privacy.
How has Data U transformed Airbnb’s culture?
Employees have taken on new roles because of their newfound data skills — something I’m always very proud of — and data has become increasingly involved in decision-making. Some decisions should not be answered solely with data, but I’m encouraged that data is being made available and is part of the process.
Your brand of data analytics closely aligns with the Andrew Young School’s new Digital Landscape Initiative, which explores the impact of new technologies on society. How would adopting Data U’s approach benefit our future as a policy school?
Data analytics can benefit any organization. The Digital Landscape Initiative addresses the ways technological advancements could affect society. In many cases, these advancements also yield new data that can be analyzed, and it’s important for organizations and governments to understand how to effectively incorporate new information into their decision-making process.