"Moneyballing" recruiting: A data-driven approach to battling bottlenecks and biases in hiring

Would you believe that your ability to hire a data scientist could be debilitated by altering just 10 innocuous words in a job description? That changing 10 words in seemingly harmless ways can result in a job taking weeks longer to fill and tens of thousands of dollars wasted?

Companies approach hiring and talent management as an art, relying on judgment and experience when conceptualizing jobs, drafting JDs, and screening and assessing candidates. With recent advances in NLP, data science and decision science, we have the ability to interrogate these “common sense” judgments to see if they help or hurt hiring teams in competitive talent markets, especially for technology and data science roles.

Maryam Jahanshahi actively studies these hiring heuristics. Join her as she showcases three behavioral studies she has recently conducted and her analyses of over 10 million jobs and their outcomes.

Topics include counterintuitive hiring patterns such as:

• How innocuous, commonly used words in job posts can significantly slow down hiring
• Why adding more requirements often lowers applicant quality instead of raising it
• Why filling jobs faster sometimes means keep the job open longer—An analysis of when qualified applicants tend to apply
• Why the most boring signals can sometimes be the most important in driving qualified applicants
• How the way we phrase things affects not only candidates but how hiring teams review applications
• Why your employee referrals could hurt instead of help the talent pipeline