Big Data Analysis: Conclusions for Recruiters

Interview: Gerhard Kenk, Crosswater Job Guide talks to Mariano Mamertino, Indeed Hiring Lab

Cross-border recruiting is based on the interlinking of international and specialised labour markets. Skills and cultural fit must match across borders to allow work environment and performance independent of a given location. Software development, medical research, aircraft engineering or foreign exchange dealing are just a few examples where these conditions prevail. Cross-border recruiting greatly enhances the flexibility of international labour markets and offer a better equalisation of supply and demand. But why would qualified experts decide to leave their own countries to endeavor abroad for better working and living conditions? When major political or economic events errupt, the pattern of job searches are a lead indicator of these trigger events. A few of these events have been covered in the international media, such as the recent US-presidential election or the decision of British voters to leave the EU. Brexit and Trump – these two trigger events have left their footprint in cross-border recruiting as a big data analysis by Indeed.com has shown. These examples are also indicative of big data analysis in recruiting.

Crosswater Job Guide talked to Mariano Mamertino of the Indeed Hiring Lab about big data analysis and the conclusions for recruiters.

 

How exactly does Indeed collect and analyse big data, especially local and regional job posts and cross border job search?

At Indeed, we have a full research unit, the Indeed Hiring Lab, dedicated to providing labour market analysis based on the data we have as the world’s largest website. Since we aggregate thousands and thousands of job boards and career websites of companies and have 200 million unique visitors each month around the world, we are uniquely qualified to comment on the global labour market.

Led by Dr. Jed Kolko, Indeed’s Chief Economist, the Hiring Lab research agenda includes large-scale labour research projects, ongoing tracking and analysis of employment trends, and surveys of industry professionals. We study the wealth of data available from Indeed in combination with government and independent research to better understand the current and future state of the job market. Combining existing data sources with the billions of job searches on Indeed, as well as the millions of resumes and employer reviews, creates a new lens on labour market trends to show how people are searching for jobs today.

 

The Hiring Lab team consists of researchers, content specialists and four economists. Actually, we are currently recruiting for further economists, including a position for DACH based in Düsseldorf, if anyone reading this interview is interested!

 

 

What statistical methodology is used to validate the results of the underlying analysis of cross border job search specifically?

Data on Indeed job searches come from a dataset of aggregated and anonymous search data from Indeed sites. We define a cross-border searcher as someone who, based on their IP address location, is currently located in one country but searches for work in another country using the Indeed site in that country. If the same searcher looks in more than one country, we count them once for each country in which they search, but we only count one search per locality per IP address in a month.

 

When you look at the data, what patterns emerge in the United Kingdom, before and after the Brexit vote?

When the Greek government held a referendum on the terms of its bailout agreement in 2015, Indeed saw a spike in Greek jobseekers searching for work elsewhere in the EU. In the aftermath of the Brexit vote, we saw a similar pattern – almost immediately, job search leaving the UK spiked and remained almost double the usual level in the days after the referendum. Countries like Ireland were the major beneficiaries of this trend, with searches from the UK for jobs in Ireland increased 2.5 times after the vote. Interestingly though, Ireland also benefited from more EU jobseeker interest, as jobseekers from the European Union previously interested in the UK starting looking elsewhere.

What is really interesting in the case of the Brexit: The spike wasn’t temporary. In fact, three months after the referendum, job searches going out of the UK were at the same elevated level as immediately after the election. Search traffic to Germany has been elevated by 9%, and Ireland continues to profit with a level 20% above pre-referendum search volume.

Does the data indicate any trends with regards to a potential “brain drain”, involving jobs for high potentials and qualified experts?

I think at this point it’s much too early to tell, any evidence towards skilled people leaving can only be anecdotal. Previous Hiring Lab research showed not only that the UK has been a net talent attractor from the EU, but that it is also by far the most popular destination for job seekers located in other EU countries. In recent years, the British labour market has been able to attract and rely on a steady flow of jobseekers and workers coming from other EU countries to fill job vacancies in the UK. But with immigration to the UK potentially getting harder after a Brexit, the challenge might not be a “brain drain”, but rather a loss of “brain gain”, as fewer EU nationals may be able to move to the UK.

And what about the US election, did you see similar trends following Trump’s win?

Yes, very similar actually – there was an immediate spike visible in our data. The share of US job searches to Canada suddenly began climbing on election night in the 9 p.m. hour (eastern time), as it became clear that Trump was going to win the election. At midnight, when the election was called for Trump, 2.7% of U.S. job searches on Indeed were for jobs in Canada — more than ten times the normal share for that time of day. A crucial difference to the Brexit data: That spike in job searches to Canada was dramatic but short lived. The share of searches to Canada was nearly back to normal by Wednesday afternoon.

 

Our data also showed that the increase in foreign searches was greater in counties that favored Clinton over Trump. To see this, we looked at the foreign search share during November 8–13 relative to the five previous weeks, for counties and metropolitan areas across the U.S., and compared it with the 2016 voting data from Dave Leip’s Atlas of U.S. Presidential Elections. The correlation between a county’s vote margin for Clinton and the increase in foreign job search share was positive and statistically significant.

But international job search is only one side of the story. Jobs in the new administration were also very popular among US jobseekers as the following infographic shows.

 

Based on the immense numbers generated by Indeed sites worldwide, the Hiring Lab also has the potential to correlate job adverts by qualification, line of industries or job titles to the demand job seekers indicate in their search. This would be a sound indication of both sides of the labour market, covering supply and demand. Do you have any plans to publish these figures?

The Hiring Lab team is actively working towards this. We are already using a metric that measures the ratio between relative employer demand and relative jobseeker interesting, for example in the Global Labour Market report we published early this year. The ratio between searches and job ads is also a good indicator that we have used, for example in an analysis of the German healthcare sector.

It is one side of the story to do number crunching, correlation calculations and more. But what is more important in the end are the conclusions drawn for the employers. How specifically will these analyses help recruiters?

The Hiring Lab specifically exists to make useful insights and information available to recruiters, we want to enrich recruiting conversations with useful data and context. It’s of immense importance to any recruiter to know which jobs are peaking in jobseeker interest and which are in decline, similarly world events prompting migration patterns could be of immediate interest. We use our real-time data and information to make recruiters and jobseeker more aware of what’s going on on the labour market. Ultimately, providing this level of insight to recruiters and jobseekers all feeds into Indeed’s mission of helping people find jobs.

 

Thank you very much for the interview.