Predictive Hiring Analytics: How it Works

-

All people leaders want to hire great people who will stay, thrive and help their organization do amazing things. However, 33% of employees leave within 90 days and 46% fail within 18 months. If heads of talent, heads of people and CEOs do not have complete data on their company’s values, culture and work environment, how can they possibly know what the “right” hire looks like and how to enable them?
 

Predictive hiring analytics describes the process of using historical data to predict which candidates will be successful employees. HR.com’s 2022 research report Envisioning the Future of Talent Acquisition found HR leaders believe predictive analytics will be the #1 most impactful emerging technology for their talent acquisition departments over the next two years.

In the Searchlight solution, predictive analytics are operationalized into talent acquisition using ‘Talent Models’, an objective talent standard based on the characteristics of successful employees. Let’s walk through what these models are, how they work, the benefits they can offer, and how organizations can build them.

Benefits of Predictive Talent Models

●    Hiring efficiently and with higher quality. Having a clear idea of your ideal candidate helps hiring teams move more quickly and make better decisions. HBR recently advised that companies look for the key one or two skills that predict success to make better hires.
●    Creating a virtuous cycle, where knowing the talent model of a successful employee equips the recruiting and hiring teams to attract more successful employees and build a high-performing culture. Tracking the skills that predict success provides objective visibility into the effectiveness of your recruiting process at bringing in the right people, and the potential to learn from and improve outcomes.
●    Reducing prestige bias and unconscious bias to identify desired candidates more accurately. More effectively hire for competencies and skills, rather than credentials.
●    Improving quality of hire to drive stronger business performance. Psychology research has found that high performers are 400% more productive than average performers.
●    Remaining competitive by improving talent intelligence at the organization and helping leadership and HR deeply understand their employee base, current skill sets, and the skillsets they need as the company evolves.

What is a Predictive Talent Model?

A predictive talent model analyzes highly successful employees at an organization and then provides a tool for recruiting to use this information to improve their hiring process. Searchlight’s Talent Models produce a holistic picture of the most successful employees by focusing on characteristics that include job-relevant skills, power skills, work motivations, culture alignment, experience, education and background. This model can then be used to inform the company’s hiring and recruiting strategy to attract more candidates likely to excel.

Since every organization is different, every model will be unique. A person that thrives at a large enterprise may struggle at an early-stage startup. There are no universal traits for highly successful employees, nor does an employee need all the characteristics in a Talent Model in order to be successful.

A successful model will compare highly successful employees with everyone else to identify what sets them apart, quantify all of these strengths and weaknesses with hard numbers, and audit for bias that could adversely impact quality candidates.

It is tempting to focus on hard metrics of employee productivity when building these models, but in the long term, the behavioral aspects of candidates are a better indicator of their success. Remember, McKinsey found that organizations in the top quartile of culture return 60% more to shareholders than median companies.

Real-world Example

I recently worked with a company as it went through the process of building and using a Predictive Talent Model for a Customer Success Manager role. Here are some details on what it did and the subsequent outcomes.

This software technology company has 500-1000 employees across the U.S. and a valuation of more than $1.5 billion. We invited current team members and two-three of their colleagues (including their direct manager) to complete reflection surveys focused on understanding their key traits.
         
The company used data from participating employees’ most recent performance reviews to identify which ones were the most successful. Then Searchlight built a unique talent profile for each person and used various data analysis techniques to surface the most prominent characteristics that differentiated the successful employees. We vetted the results to check against adverse impact and found no evidence of bias, meaning that the analysis was not flagging certain genders or ethnicities as high or low performers at significant levels.  

They found that the most successful employees at their company (determined by their performance reviews) tended to be fast learners, results-oriented, resourceful, intuitive, and more likely to be strategic thinkers. They were also more likely to enjoy working on a variety of challenges, rather than diving deep into one area, and prioritized speed over quality.

You can see how these strengths might be different in another organization or role. Some jobs might require more of a focus on quality over speed, or reward specialization. This is why predictive talent models are so helpful - they show what your specific organization needs from its employees.    

This company updated some of its recruiting and hiring practices based on the Talent Model and saw excellent results. Time to Productivity for Customer Success new hires decreased by 25%, and from 4 months to 3 months. Anecdotal feedback from hiring managers was that the new hire quality was stronger than before. This better onboarding experience and shorter ramp time had a positive impact on morale, culture, teamwork and other key elements of productivity.

This company estimated that hiring a top performer and decreasing ramp time improves Employee Lifetime Value by 1.5-4x. This lines up with research out of Indiana University that high performers can deliver up to 400% more productivity than the average employee.     

All in all, Predictive Talent Models offer a fantastic way to boost hiring speed and efficiency, improve quality of hire, and generate business value in many ways. I strongly urge forward-looking CEOs and people leaders to consider implementing them.

For more information about Predictive Talent Models and more details from real-world deployments, read Searchlight’s recent white paper.

Author Bio

Anna Wang is CTO and Co-Founder of Searchlight. Anna graduated from Stanford with a B.S. in Computer Science and an M.S. in Artificial Intelligence. Prior to Searchlight, she worked as a Product Manager at Google, software engineer at Uber, and management consultant at McKinsey. She is a Forbes 2021 30 under 30 honoree. Visit searchlight.ai Connect Anna X. Wang