By Adam Hyder, Chief Technology Officer, Weijen Hsu, Senior Software Engineer, and Chet Kumar, Big Data Architect
At Jobvite, we employ cutting edge analytics techniques to mine recruiting intelligence from our comprehensive database of over 50M job seekers. Combine that with over 10 years of behavioral recruiting data, it’s clear that we are sitting on a gold mine; it allows us to slice and dice the data to garner fresh insights about job seekers, companies, industries, and greater economic and hiring trends.
After working with thousands of customers and mining our database, we have a unique perspective on what factors lead to successful hires because we can track candidates as they progress through the entire recruiting funnel–from how they are sourced, hired, and onboarded.
We’re an innately curious bunch, so we’re always posing questions that might uncover interesting insights and then diving into the data to see what we find. For example, we wanted to better understand the impact of the total number of candidates for a particular job and how it impacts time-to-fill (TTF). At first blush, it would seem that a larger number of candidates is always a good thing. However, when we ran correlations on ‘job fill success rate’ vs number of candidates, we found that the correlation coefficient R was actually negative, -1 <= R < 0.
Maybe too many candidates for a position isn’t a good thing after all. What if you have 1,000 applications for one job? How can you possibly look at 1,000 resumes without quickly becoming overwhelmed?
This got us thinking. Is there a certain number that divides the line between optimal and negative? We wanted to find out. We approached this problem by slicing our data into 31 different job categories–such as Engineering, IT, Sales, Marketing, etc.–which we are further refining to create a comprehensive taxonomy of jobs types.
After that, we found a sweet spot in the number of candidates that maximizes the likelihood of a successful and timely hire. Anything less, or more, than this range negatively impacts likelihood to fill a position. This sweet spot even varies by the job category.
For example, for Engineering jobs, the sweet spot range is between 20-50 candidates to maximize probability to hire, shown in chart below. We are utilizing a similar approach to develop Industry and Company benchmarks on TTF that will provide Jobvite customers useful metrics about how to improve their recruiting performance.