LinkedIn Ran Secret Test On 20M Users To Rate Job Success From Their Connections

Over the last five years, professional networking site LinkedIn has been running an experiment on 20 million users to analyze the success rate of job-hunting.

The results showed that one is more likely to secure new jobs via ‘weaker ties’ than through people close to them. 

The study, published in the Science journal, was conducted by Harvard Business and the Massachusetts Institute of Technology (MIT). LinkedIn’s own set of researchers also joined the schools. The project aimed to better attune job recommendations and mobility across the platform. 

To gain data for the study, several “large-scale randomized experiments” were carried out on the platform’s “People You May Know” tab. This tab links users up with people they might potentially know. 

Using A/B testing, the experiment provided people with different contact recommendations, and then it examined the work offers produced from these 2 billion new connections. Out of those links came 70 million applications, leading to 600,000 new jobs.

The experiment was based on the “strength of weak ties” theory. According to Mark Garvenotter, a professor at Stanford, the people we know can be classified into two groups. Relatively “weak contacts” share about 10 connections, and “stronger” ties have 20 or more relationships in common, New York Times detailed.

A year into the study, the researchers found that people with weak connections were likelier to land work in companies with those acquaintances than those in their “strong tie” network. 

Though it should be noted that the results did vary between industries, more digital-based industries thrived better in their weak ties network, while sectors that relied less on software systems found jobs more through their strong ties. 

Professor Aral from MIT, one of the authors, said its primary purpose was to prove how vital social networking platforms can be and highlight significant economic issues. 

While the findings surfaced useful prospects, the lack of transparency in the research process has raised some eyebrows. 

For some, what was the most revealing discovery was that the tests were conducted behind users’ backs.

LinkedIn’s privacy policy, however, states that user data can or will be used in research experiments. Furthermore, a statement from the company mentions how the research techniques were “non-invasive.” 

As more corporations come under heat for surveillance of users without consent, privacy concerns are often at the center of debates. 

Weak ties or strong ones, some things are best kept private.

[via USA Today and The New York Times, Photo 164725791 © Andreistanescu | Dreamstime.com]