HiringBranch Study: Scoring Soft Skills Separately Weakens Frontline Hiring Predictions

Frequently Asked Questions

What did HiringBranch's new research actually find about scoring soft skills?

HiringBranch compared machine scores for individual skills like empathy and acknowledgment against human annotator ratings and found only moderate correlation when skills were scored separately. When the same skills were grouped into a proprietary machine learning model with different weights and calibrations, correlations with human annotators became much stronger, suggesting combined evaluation predicts frontline soft skills more reliably than isolated scores.

Who is Assaf Bar-Moshe and what does HiringBranch do?

Assaf Bar-Moshe is Chief of Research and Development Officer at HiringBranch, a Montreal-based assessment company. HiringBranch builds entirely open-ended assessments for frontline roles such as customer service, sales agents, and retail associates, placing candidates in scenarios where they respond by voice or in writing rather than choosing from multiple-choice options.

Why does HiringBranch use a linguistic approach instead of personality testing?

Bar-Moshe, a trained linguist, explained that HiringBranch performs a "sociopragmatic analysis of the words that the candidate is actually saying" to determine whether responses are acceptable in a given situation. This differs from personality tests because it anchors every skill, including emerging ones like integrity, grit, creativity, and adaptability, to how it is actually expressed through language on the job.

What are the four pillars of customer service HiringBranch measures?

HiringBranch built its customer service study around four prominent skills: acknowledging the customer's issue, reassuring the customer using positive language, empathy, and active listening. Bar-Moshe emphasized these skills do not operate in isolation, noting that empathy without the ability to comprehend and solve the issue is "nice, but it's actually useless."

How does HiringBranch turn a job description into an assessment?

HiringBranch translates job descriptions into conversation flows, the replicable back-and-forth patterns a candidate would handle daily, such as listening to a complaint, reiterating understanding, and offering a solution. Those flows become scenarios that elicit specific skills. When descriptions are weak, the team supplements with online research, recruiter conversations, and actual customer conversation transcripts provided by clients.

How are assessments calibrated and validated over time?

Calibration is client-specific and can vary by industry, role, and even geography, with different weights for markets like Vancouver, Toronto, and Montreal. HiringBranch also loops back post-hire, comparing assessment performance on quality and customer focus with actual job performance months later, and feeds those findings back into the models so assessments get smarter over time.

What new self-serve capability did Bar-Moshe preview?

Bar-Moshe confirmed HiringBranch is actively working on a self-serve tool that would let hiring managers build their own assessments using a library of conversation flows and skills drawn from similar roles. The goal is to help hiring managers, who often know they need to fill a job but not the job itself, start from guided templates rather than a blank screen.