Executive Summary

In late 2017, a global sports apparel brand found itself in a frustrating loop: The company had rapidly hired store managers and sales associates for their Toronto-area stores, but many of the hires proved disappointing, underperforming and leading to poor sales for the seven regional stores. This led to terminations, which then led to more rapid hiring.

To put an end to this costly cycle, management partnered with Empirical Hire with the goal of continuing to move quickly while making sure the RIGHT managers and sales associates were hired. The results:

The turnaround was due to a significant shift in thinking on the part of management: an understanding that the key was finding the RIGHT people for the company. They needed store managers and sales associates whose experience, but also soft skills and motivation levels, would lead to strong performance.

Empirical, whose technology is powered by behavioral science, was able to build a complete understanding of each candidate, identify what constitutes success for each role at the company, and then make the matches, finding the right candidates for each position. The hiring remained rapid, but now it was also targeted and accurate, leading to higher sales and increased retention.

‍

Business Challenge

In late 2017, the management team was tasked with setting up 7 stores across a major metropolitan area in a 12-month period, essentially building the entire workforce from scratch. The team went “all in” and immediately hired store managers. The newly-hired store managers, in turn, were tasked with quickly staffing their stores with sales associates. Candidates were sourced from online job boards, advertisements in retail malls, and job fairs. Because store managers screened candidates and conducted interviews, they were able to rapidly fully staff the new stores.

A few weeks in, store managers started expressing their dissatisfaction with some of the employees; employees showed up late or called in sick, and those who did come to work were not sufficiently sales-driven and were easily demotivated by the need to work harder and cover for those who did not show. In order to maintain the supreme level of service this brand is known for, the Management team decided to act decisively and let go of underperforming employees. They took the same approach with store managers that failed to motivate and lead their team and create a culture of excellence.

In order to keep up with business, the remaining store managers were instructed to hire in volume, and so they did. This policy quickly resulted in overstaffing of stores while simultaneously leading to voluntary and involuntary turnover of employees, mainly sales associates directly responsible for sales and customer service. Turnover rates remained high and were growing from week to week.  At that point in time, the management team came to the realization that something needed to be fundamentally improved in their hiring processes.

Understanding The Problems

The Head of Sales approached Empirical with a list of challenges

1. Hiring in high volume compromised quality - Hiring managers’ main hiring KPI was time-to-fill with no connection to quality of hire, which led to quick, superficial interviews and rushed hiring decisions. There was a detailed job description, but managers highly differed in their perception of what makes a good fit for a store associate in terms of personality traits, skills, behaviors, and bio (i.e., experience, education, certifications, etc.).
2. Lack of standardization in the hiring process - Process structure and methodology highly differed between store managers. Some managers conducted phone interviews, some did not, some conducted long interviews, some short, and each manager asked different questions in the interviewing process.
3. Heavy reliance on intuition rather than data - At the end of the interviewing process, managers described making decisions mainly based on their gut, and “knowing in the first 5 minutes if the candidate is a fit or not.”rather than justifying their decisions with data.

‍

Hiring with Empirical

Empirical’s goal was to create a quality-driven, structured, bias-free hiring process that would ultimately reduce turnover and increase sales.

1. Profile position - Together with the store managers, Empirical’s Customer Success team, all I/O psychologists, defined the core traits, skills, behaviors and biographical data points (i.e., experience, education, certifications, etc.) that a great store associate possesses. The profile enabled greater clarity and alignment amongst hiring managers with regards to what they should look for in candidates.
2. Define job KPIs for successful sales associates - Empirical and the management defined outcome measures in order to prove Empirical’s impact, in terms of retention and sales performance.
3. Measure improvement - Exactly like in scientific experiments, Management requested that we demonstrate Empirical’s impact by comparing performance of sales associates hired with Empirical’s technology and those that were not. Store managers randomly assigned candidates to be screened and hired with the help of Empirical’s technology or by the current hiring process.
4. Build a measurable hiring process - Empirical’s online assessment assessed the traits and skills defined in the profile. In addition, Empirical integrated  questions that were found to predict performance in other similar retail organizations. Most importantly, Empirical’s Interview platform provided one central place for store managers and management to clearly track every stage of the hiring process. Because all of the candidate data is automatically aggregated in one, clear, intuitive mobile and tablet-friendly application, store managers could make data-driven decisions because the data was available at their fingertips, rather than buried in paper applications, random notepads, or their memory.
   
   In order to amplify assessment accuracy, Empirical also gave it to all existing store associates, and  provided Empirical with each associate’s performance KPIs. By analyzing assessment results with actual job performance Empirical’s AI was able to extract which profile traits actually predict performance. The result was a leaner, more focused profile that allowed store managers to focus on gathering information only regarding the predictive traits during the recruiting process.
   In addition to the online assessment, Empirical’s platform includes a structured online interview tool, allowing managers to:  

• Ask each candidate questions that are the most relevant for their points of strength and weakness as arose from the assessment,
• Score interview questions and compare candidates based on their final score, and
• Rely on information from behavioral questions rather than just gut feel.

5. Apply Empirical’s predictive machine learning models - Having a measurable hiring process in place, Empirical used data collected,along with sales and retention data, to train machine learning models. In this process, Empirical learns how different personality correlate to employee performance. The algorithms learn from:

• Environmental normalization of employee’s performance score
• Analysis of test answers and response patterns
• ‘Leveling’ of pre and post calibration answers using singular value decomposition (SVD)
• Two Phased model training: bootstrapping at ~1000 iterations and evaluating the training quality score to ensure that models are unbiased
• Final evaluation using untrained data

6. Continuous AI-driven improvement - The power of Empirical’s technology is that the algorithms continuously enhance predictive components while simultaneously omitting non-predictive elements. The result is increased hiring accuracy over time.

‍