How LinkedIn Data Can Make Our Workplaces Better

Big data can help companies pinpoint potential revenue, create targeted marketing campaigns, and improve customer service. What might data also reveal about gender equality and inclusion? And how can we apply that data to help diversify pipelines and recruit more women into male-dominated industries?
LinkedIn’s vast data may provide some answers: their Economic Graph maps the professional platform’s data from 590 million members, 50 thousand skills, 30 million companies, 20 million open jobs, and 84 thousand schools to spot workplace, job, and economic trends.
Allen Blue, VP of Product Management and Co-Founder of LinkedIn, and Nikki Waller, Editor, Live Journalism and Special Coverage at The Wall Street Journal sat down in The FQ Lounge, Home of Equality @ Davos to discuss how data, analytics, and inclusivity fit together. You can watch the Q&A session, and read the key insights below.

Allen Blue of LinkedIn and Nikkie Waller of the WSJ in The FQ Lounge @ Davos

Allen Blue of LinkedIn and Nikki Waller of the WSJ in The FQ Lounge @ Davos

Nikki: What can LinkedIn’s job-matching data show us about equality, diversity, and inclusion?

Allen: When we started LinkedIn, it was all about people — and people connecting with one each other. But on their LinkedIn profiles, people also put companies they worked at and schools they went to. So it turns out that LinkedIn represents more than just people: it’s about companies, schools, and more. So, when we took a step back, these patterns allow us to see how the economy works, especially from an HR perspective. When it comes to talent, to knowing where the skill gaps are, to knowing which skills are in demand or surplus – you can figure that out through LinkedIn.
LinkedIn collects data on gender patterns.  Bias comes in a lot of different ways, and it often plays into hiring. So we have a proxy for that through LinkedIn. Our system looks at people’s first names, which is surprisingly accurate, and from that, we can see how men and women behave differently in the job search process and in their self-representation.

Nikki: What are some of the most powerful data sets that businesses can tap into to advance inclusion?

Allen: Take advantage of any data they can look at. Driving equality isn’t just the right thing to do: it’s the effective thing to do. So for a hiring manager, it’s about looking at all of the people who might be able to fill a role, which means that your pipeline is way larger.

Nikki: If I’m a manager at LinkedIn, how is the push for inclusion being communicated to me?

Executive involvement is key. It’s very hard for the hiring team to solely drive inclusion if the executive team isn’t actively behind the process. We have a very transparent path from CEO and executive team to the rest of the employees – and we routinely communicate to all of our staff about our inclusion initiatives.

Outside The FQ Lounge, Home of Equality @ Davos

Outside The FQ Lounge, Home of Equality @ Davos

Nikki: Women comprise just 22% of the AI workforce. How can we get more women in AI?

Allen: Yes. When we looked at all the people on LinkedIn who have artificial intelligence skills as part of their profile, just 22% are women.
The thing I would say is most hopeful about bringing women into AI is that AI and machine learning are topics of interest to virtually every industry in the world. And anyone with AI skills can be part of that conversation. From speaking with female AI professionals who work at LinkedIn, people with STEM and biology backgrounds have a lot of similar experiences that a person needs to do AI well. So there’s a very forward path for a woman who graduated with a Master’s degree or Ph.D. in biology to get into the world of AI.
Having a diverse group of people building these tools isn’t a complete solution, but if we’re going to build a full set of expectations of what AI can do for us, we need a diverse group of people actually building those tools.

Nikki: LinkedIn is working with a pretty incredible set of data. How can that data be used to nudge recruiters whose hiring is a little off-balance?

Allen: The first thing we need to do is make sure the machines are actually providing the best possible data and results. An example: LinkedIn recently went in to make sure that the data we provide about candidates isn’t biased – that men don’t fill up the first pages of searches in lieu of qualified women. That’s step one.
Step two is to introduce new capabilities. So we’ve begun to help recruiters move away from things that are pure recognition — like where a candidate went to school — to what skills the candidate has. That transition is an important one. A lot of recruiters make decisions based on where someone went to school. But really they’re just looking for the best person for the job. In the end, the key thing is not about having machines solve all the problems. It’s really about the human-machine combination. We can do a perfect job at LinkedIn, but those biases will still exist when the candidate gets into the hiring process at the company. That’s why we want to work with recruiters to make sure they know what tools are available to them to eliminate bias once they take over in the hiring process.
Harnessing data to drive more inclusion in the workplace can be a powerful tool, but it’s up to humans to use that data to take positive action steps for change. For more leaders in The FQ Lounge @ Davos who are taking giant steps forward for equality, check out:
What Happened in The FQ Lounge, Home of Equality @ Davos: Day One
What Happened in The FQ Lounge @ Davos: Day Two
Why Diversity Should Be a Business Goal