LinkedIn is widely regarded as a social network for grownups, connecting 259 million people worldwide who put their résumés on display. It was never intended to create a teen haven. While parents may enjoy posting multidecade work histories, it’s harder to get high-school-age babysitters and burger flippers excited about documenting part-time jobs.
Enter the data scientists. As early as 2011, LinkedIn began rethinking how it wanted to interact with the under-18 set. Teens might not have much to contribute to LinkedIn’s 20-petabyte trove of career information, but they could become some of its most avid data consumers. Specifically, LinkedIn could build a way for them to see where alumni of specific colleges end up working—giving teens a kind of analytics dashboard to lay odds on their futures.
Take well-known U.S. universities such as Carnegie Mellon and Purdue. In each case, LinkedIn has data on the career paths of more than 60,000 graduates. That’s a data set big enough to allow for some fascinating fine-grained distinctions. Type in MIT, and you quickly learn that graduates are unusually likely to land jobs at Google, IBM, and Oracle. Plug in Purdue, and employers such as Lilly, Cummins, and Boeing predominate.
Such information is a gold mine for high-school juniors and seniors, says Purvi Modi, a college advisor in Cupertino, California, since most high-school students have only a hazy idea of what careers are out there. By using LinkedIn’s tool, students interested in specialties such as solar energy, screenwriting, or making medical devices can pinpoint schools with the best track records of sending graduates into those fields. Modi, who advises about 300 students a year, says about 40 percent of them now cruise through this part of LinkedIn’s database, known as University Pages, to get insights. That’s impressive, given that the data-combing service has been fully available only since August 2013.
LinkedIn makes money from its huge enrollment in two ways. Recruiters pay as much as $8,500 a year for enhanced access to job candidates, while members can buy various premium services that make it easier to navigate the site. Investors think LinkedIn could be creating a near-monopoly in the global market for talent. As of January, the company was valued at $24.5 billion (a remarkable 728 times annual earnings), reflecting a belief that the social network has just begun to tap the value of its enormous databases.
That high valuation also puts pressure on LinkedIn’s team of 68 data scientists to build new tools to extract value from all those petabytes. One set of algorithms now coaches recruiters on “People You May Want to Hire.” Other tools alert restless workers to “Jobs You Might Like.” The University Pages initiative fits into this pattern; it’s a low-key version of “Colleges You Might Want to Attend.”
Creating the right tool for college hunters turned out to be surprisingly tricky, says LinkedIn’s lead data scientist on the project, Gloria Lau. There wasn’t any good way to offer instant lists of good college choices, because teens (along with their parents) generally lack clarity about their priorities at first.
Young people need time to explore on their own, Lau found. By playing with various filters, students who start with broad interests in an area such as engineering can discover subdisciplines and employers that they mightn’t have known about at first. One teen might end up being interested in mechanical engineering careers at Tesla or Lockheed Martin, while another might learn that the local college is likely to lead to jobs in petroleum engineering at Halliburton.
This slice-your-own-data approach is slower and more unpredictable than instant recommendations, like those LinkedIn gives to job seekers. From LinkedIn’s perspective, that’s not all bad. College hunters tend to linger on the site, so they may see more ads or be drawn into greater use of LinkedIn. By letting users shape their own requests, LinkedIn also avoids playing favorites or issuing unflattering reports about specific schools.
So far, LinkedIn isn’t charging either student users or campuses for any of University Pages’ features. But even a free service can help business objectives. The obvious payoff, says data chief Jim Baer, can be seen in new-member data. LinkedIn’s membership is growing by 38 percent a year, with the fastest expansion in the student and recent-graduate segment.
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist
An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.
The Biggest Questions: What is death?
New neuroscience is challenging our understanding of the dying process—bringing opportunities for the living.
How to fix the internet
If we want online discourse to improve, we need to move beyond the big platforms.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.