4 Ways to Bring Big Data Into Your Recruiting Process
Welcome to Recruiter Q&A, where we pose employment-related questions to the experts and share their answers! Have a question on jobs data you’d like to ask? Leave it in the comments, and you might just see it in the next installment of Recruiter Q&A!
Today’s Question: We’re looking for real-life examples of big data in action. How do you use it in your recruiting process? What results have you seen?
1. Determine Who Your Best Hires Really Are
In talent acquisition, we often source candidates based on ideal profiles. It’s quite common for hiring managers and recruiters to create lists of qualifiers and preferred candidate characteristics and background experiences that they perceive to be critical to success in the employer’s work environment and culture. Big data is a great way to evaluate these ideal profiles.
For example, see our work recruiting thousands of college interns. Many of the hiring mangers we were supporting strongly believed that college students from highly competitive universities were better matches for their internship programs. As a result, the majority of the recruiting budget and resources were directed to the most competitive schools, which included the Ivy League. When we examined internship evaluations on a school-year basis, we saw great results. The number of hires needed was achieved, and hiring managers rated the internship pool as highly qualified. The performance of interns based on hiring manager reviews also exceeded the established targets.
But as we entered year three of the program, we noticed that a similar number of hires was consistently being requested. So, we looked into how those interns from year one and year two were doing in year three. As it turned out, most of those candidates were no longer with the employer. Although their job performances were rated strong, candidates from the most competitive programs either resigned or did not accept the employer’s offer for a regular, full-time position upon graduation.
When we combined two years of data, which included thousands of screened, interviewed, selected, and employed candidates, we observed that students with the highest GPAs from less competitive schools were twice as likely to become full-time employees and build careers with the employer than those from the top universities. The data set also showed that these students performed equivalently to those from the more competitive schools, who tended to pursue careers in other industries or with other types of employers.
2. Anticipate Your Candidates’ Questions and Engage Them From the Start
At LifeGuides, we use more than 3 million data points to tell us what candidates care about, and then we tailor employer branding content toward whatever gets the most engagement. Thanks to these efforts, our videos have twice the level of engagement as other videos.
Here’s an example on something we put together to give sales candidates a better understanding of our company and the role. We decided upon which questions to address in this overview after looking at the data regarding which questions candidates engage with the most when it comes to content from other companies on LifeGuides.
3. Tie Candidate Assessments to Meaningful Performance Metrics
I have seen the use of assessments in selection later tied to performance metrics, and regression analysis can be run to see which parts of the assessment have the best predictive qualities to performance. You need to collect data properly on the front end in order to have predictive abilities.
Assessments used for selection should continue to be data points used when an employee is hired for performance management. With a large enough data set, you can see which specifics of an assessment will have the highest predictive components for driving performance. For example, if you measure 10 different things in selection, over time, you might see that three of them show the biggest common drivers of high-performing employees. This will allow you to focus on making decisions with the highest predictive abilities.
At Scout, a recruitment marketplace, we use big data to determine recruiters’ interests in working a job and their likelihood of success in filling the job. That way, we can match the right recruiter with each job listing. The data is based on billions of data points related to historical activity and past performance in filling roles. Data points include responsiveness, job type, industry, geography, relationship, current activity level, and fee potential.
