In the latest episode of the podcast You Should Know, hosted by William Tincup, Assaf Bar-Moshe, Chief of Research and Development Officer at HiringBranch, discussed new research that challenges a common practice in pre-hire evaluations: reporting soft skills such as empathy, acknowledgment, active listening, and reassurance as isolated scores. The episode, titled "Assessing Skills One at a Time Is Costing You Better Hires," was published on August 26, 2026, and comes as employers increasingly focus on measuring frontline talent—customer service reps, sales agents, retail associates—against the realities of live customer interactions.
Bar-Moshe explained that HiringBranch's open-ended, voice-and-writing assessment design was used in the study, which found that single-skill scoring produces only moderate correlation with human annotators, while a combined proprietary model yields much stronger correlations. The research underscores the importance of evaluating soft skills in tandem rather than in isolation, because real-world customer interactions require a blend of abilities.
The four pillars of customer service that HiringBranch measures are acknowledgment, reassurance through positive language, empathy, and active listening. Bar-Moshe emphasized that these skills are not independent but interdependent. "If a candidate can express empathy, but is unable to solve the issue correctly or to comprehend the issue correctly or to reassure the customer, then this empathy is nice, but it's actually useless," he said. This highlights why a holistic assessment is more predictive of on-the-job performance.
The discussion also covered how job descriptions are translated into conversation flows and scenario-based assessments calibrated per client, region, and role. HiringBranch takes a linguistic rather than personality-based approach, described by Bar-Moshe as a "sociopragmatic analysis of the words that the candidate is actually saying." The company's team of IO psychologists and linguists uses years of textual data to build machine learning models that predict empathy, acknowledgment, and related skills, then validates those predictions against on-the-job performance months after hire.
Tincup recounted a retail confrontation over mispriced broccoli that turned on diplomacy rather than policy, illustrating the complexity of real-world customer service. Bar-Moshe used this example to show how a candidate might demonstrate empathy but fail to resolve the issue, underscoring the need for a combined scoring model.
The research also found that calibrations differ by client, industry, and geography. Regional variation across markets like Vancouver, Toronto, and Montreal shapes different scoring weights for the same role. Bar-Moshe also previewed a self-serve capability in development that would let hiring managers build assessments from a library of conversation flows and skills, reducing reliance on weak or generic job descriptions. The full study will appear under the AI research tab on the HiringBranch website.
This research has significant implications for hiring practices. By moving away from isolated skill scores, employers can improve the predictive validity of their assessments and make better hiring decisions for frontline roles. As the workforce evolves, the need for comprehensive evaluation methods becomes more critical, and studies like this provide a data-driven path forward.
