Inside LinkedIn’s quest for talent insight
Nearly since its founding, LinkedIn had been sitting on a vast trove of data about its users, their job skills and job moves. Today the Microsoft subsidiary has around 546 million users in 200 countries and territories. In the third quarter of 2018, LinkedIn plans to release “Talent Insights,” a resource intended to give HR executives and recruiters insights that can help them both strategically and tactically. Talent Insights was announced last October, and a cross-section of invited “charter” employers, both large and small, has already been using the product on a trial basis for several months.
To learn about LinkedIn’s plans for Talent Insights and its capabilities, EBN recently spoke to Eric Owski, LinkedIn’s head of product, talent insights and talent brand. Edited excerpts of that conversation follow.
Employee Benefit News: What’s the general idea behind Talent Insights?
Eric Owski: Over last several years it’s become clear to us that the talent acquisition function has been under a lot of pressure to make decisions using data. As the war for talent has become more intense, the decisions about site location and long-term workforce planning needs to be built in a solid foundation of data and insight. We have a phenomenal amount of data on what’s happening in the labor markets globally, some of it happening in real time, so it made sense for us to try to meet that need.
EBN: In general, what kind of data are you talking about?
Owski: For starters, what companies are growing which skill sets? How are people moving between companies all over the world? [Those are] things that are captured on our platform.
EBN: OK, so let’s say I want to know about what’s going on in the world of people with skills in machine learning. What can Talent Insights tell me?
Owski: First, the system will produce a map showing you the cities that have the most people with that skill, in the U.S. or globally. Then you can learn how many have changed jobs in last year, how long they stay in a position before they change jobs, among other things.
EBN: So all of this is just gleaned from data LinkedIn users post themselves? Are there any privacy concerns?
Owski: The data we’re showing in Talent Insights is aggregate data. And below a certain threshold, we won’t expose data for the reason that we want to protect user privacy. Member privacy is a priority for us. So if people don’t even want to be counted in Talent Insights, they can opt out of that. It’s in their privacy settings.
EBN: OK, what else can people learn from the data?
Owski: A lot more. For example, you can learn the level of demand for people with particular talents, based on job listings and how many in-mails are sent to this population. You can learn who are the top employers of people with particular skills in each market.
EBN: What about the tightest labor markets for people with whatever skill you’re looking for?
Owski: Talent Solutions will show you the demand for a particular category of talent relative to the supply. So, for example, you can use that to find “hidden gem” locations from a recruiter’s perspective — places where there’s low hiring demand relative to the supply there. It could be cities like Raleigh-Durham, Houston, Detroit and Phoenix. These might be good markets if companies are willing to be flexible in making their location decisions.
EBN: I assume those places change over time.
Owski: Right. We’ve just introduced a new module that shows talent migration patterns over last 12 months — the net gain and loss of talent by geographic location. It can show you how many people have moved in and out of a particular market, where they’re going and where they have come from.
EBN: Everybody doesn’t use the same precise terminology to describe the same skills, correct? Does that muddy the data?
Owski: To be sure you capture all the people you’re looking for, in a search you can add other terms, like supplementing “machine learning” with “artificial intelligence.”
EBN: So far we’ve been talking about people working in high tech. What about lower tech jobs like machine tool operator?
Owski: LinkedIn has traditionally been a tool for professionals, but we do have millions of members who are not in professional categories. Still, I think the companies that are going to use this tool over the next few years are going to have the most success in those professional categories, like engineering, finance technology, research and science. Those are the most common searches we are seeing on the platform now. As we get more data, companies will be able to use the tool for those other categories as well.
EBN: Can talk more about what kind of company-specific data is available?
Owski: We’re able to show which the top employers are by job category, and how fast that part of their workforce is growing, and compare that to other companies. And we’ve just introduced the ability to measure attrition rates — the gains and losses of talent that a company has by talent category. So you can see which companies have the lowest attrition rate for machine learning talent. That’s measured by the number of departures in proportion to the average number of employees at that company within that job category.
EBN: So I guess that’s a pretty good way for companies to benchmark themselves against the competition, whether they’re going a good job of retaining certain kinds of talent.
Owski: Exactly. And you can also learn how strong your brand within a particular talent pool. So if I’m a recruiter and send an in-mail to candidates, what is the response rate to my in-mails compared to a competitor? Soon we’ll have a benchmark that will show how well companies compare to their competitors in that respect. We’ll also have results from our global talent-driver survey, a brand perception survey where we send polls to hundreds of thousands of job-seekers and ask how they perceive the various employment brand attributes of particular companies.
EBN: Did you say you can show where companies employ particular categories of talent?
Owski: Yes, as you go deeper, you can see how their workforce is distributed across the world, where their pockets of talent are, and where are they growing the fastest. Once you get into a company report, all the data is benchmarked against your own data. So for example you look at a company that has 29% of its workforce with a specific skill set in a particular city, and your own company’s might be zero.
EBN: So if I’m trying to hire people with particular skills at a particular location, I could use this tool to raid a company there that has lots of those people?
Owski: We are finding companies in our charter program that are using this data competitively, but there is a broader use case beyond talent acquisition. The workforce composition benchmarks we provide will be very useful for workforce analytics for HR professionals who are thinking about long-term workforce planning, and how to grow their workforce in a way to set themselves up for future success.
EBN: How do your account for gig workers?
Owski: We’re doing our best to distinguish between people who are full-time workers, and others who are contract workers. Later this year we will add a filter for “employment type,” either using explicit data provided by members that we have, or by inference.
EBN: Does Talent Insights provide salary data to help employers determine what they’d have to pay for a particular category of talent in a given geographic location?
Owski: That’s one of the things we want to roll out over the next couple of quarters.
EBN: How does Talent Insights integrate with LinkedIn Recruiter?
Owski: In two ways: You’ll be able to imbed Talent Insights into your recruitment system. It’s not realistic to think a company is going to adopt a completely new analytics data while they’re trying to manage their requisition load. And for recruiters who are going to be primarily living within Talent Insights, any query you to do can be translated into a LinkedIn Recruiter search. So if you’re doing an aggregate analysis of the labor market, and you have a particular query that you like and you’re licensed for both products, you’ll be able to hop right over to Linked In Recruiter and take action on the candidates who are on that list. So the integrations will be pretty tight.
EBN: Does one need to be a very large employer to get the full benefit of this resource?
Owski: In our charter program we have been using companies that represent a cross-section of our customer base, from multinational companies to much smaller ones. All sizes of companies that are consistently hiring are thinking about strategic questions. They might be running into talent supply limitations in the markets they are in, so they’re thinking about opening new locations, or deciding to hire scarce talent to work remotely. I don’t think there’s a minimum size. As soon as talent becomes a strategic priority for them, even if they only have 50 employees, this product can answer questions that are very essential for them to scale successfully.