Airbnb’s Data University Aims to Address a Critical Skills Gap

Airbnb’s Data University Aims to Address a Critical Skills Gap

In the era of big data and analytics, data scientists are hot property in the labor market right now: A report released last month by the Business-Higher Education Forum and PwC warned that the US is facing a huge skills gap in this field, with 69 percent of employers who say they need candidates with data science and analytics skills, compared to just 23 percent of college and university leaders who say their students are graduating with those skills.

An analysis of data from the Bureau of Labor Statistics earlier this year showed that data scientists were at the top of the list of the most in-demand professions, and according to Glassdoor, they are earning some of the highest salaries in the country. Yet the US, it seems, is simply not producing enough data science experts to fill the large and growing demand.

One way to address this gap is to invest more in university data science programs, but employers don’t have the time to wait for new classes of graduates to rise through the college pipeline. With that in mind, John Mannes reports at TechCrunch, Airbnb is trying a different approach—introducing its own internal, university-style data science program, “complete with a custom course-numbering system”:

Data University is Airbnb’s attempt to make its entire workforce more data literate. Traditional online programs like Coursera and Udacity just weren’t getting the job done because they were not tailored to Airbnb’s internal data and tools. So the company decided to design a bunch of courses of its own around three levels of instruction for different employee needs.

100-level classes on data-informed decision making have been designed to be applicable to all teams, including human resources and business development. Middle-tier classes on SQL and Superset have enabled some non-technical employees to take on roles as project managers, and more intensive courses on Python and machine learning have helped engineers brush up on necessary skills for projects.

Since launching the program in Q3 2016, Airbnb has seen the weekly active users of its internal data science tools rise from 30 to 45 percent. A total of 500 Airbnb employees have taken at least one class — and Airbnb has yet to expand the program to all 22 of its offices.

I think this is a great approach, as it works on several levels. The basic quantitative skills taught in the “100-level” classes are going to be useful pretty much across the board, and making the more advanced content available may enable Airbnb to source a very valuable, specific skill set. And there’s no reason why that shouldn’t be achievable. Data scientists are often hired externally from an academic background, but they didn’t spend five years in their PhD program learning how to use Python; if anything, those skills were almost a side-effect of doing their actual research. It’s not easy to learn that stuff, but it’s very attainable for people who have the aptitude.

Just as we’ve seen more and more organizations opening up to hiring programmers out of coding bootcamps, I wouldn’t be surprised if more employers started trying this internal approach to data skill building. We tend to think of a “data scientist” as a type of person, but it’s really not: It’s just someone who has had exposure to the right learning opportunities, and there’s no reason progressive organizations can’t provide that themselves.