The Candidates You Will Never See
- David Frank
- Jan 20
- 6 min read
This subject pressed on me after speaking with a colleague who described watching highly capable candidates disappear in the first round of automated screening. They never failed an interview, never stumbled over a question, never faltered in presentation, they simply never made it through the filter. It left me unsettled. If talent is present, yet invisible, what does that absence cost both sides? It seems urgent to ask because these filtering systems are not going away. The future of hiring is not less automation, but more, which means the only path forward is better calibration.
The Filtered Future
Hiring has become a contest not between people, but between patterns. Automated tracking systems, resume parsers, AI screeners, online assessments, and background-check algorithms form a mesh of filters. Each was designed to create efficiency at scale. Yet efficiency conceals omission. Research from Harvard Business School shows that 88 percent of qualified candidates are filtered out by applicant tracking systems due to minor formatting issues or missing keywords.¹
This is not always incompetence. Many of those removed are fully qualified, even exceptional. Their rejection comes from criteria set long before they applied: a missing keyword, an employment gap, an unconventional degree. What employers see as precision is often just reduction. For employees, the barrier is not merit but invisibility.
And yet here lies potential: recalibrated correctly, these same tools can widen access, mitigate bias, and surface talent that traditional processes overlooked. Employers gain richer pipelines, employees gain pathways to demonstrate ability beyond rigid proxies.
Qualified but Discarded
The paradox is that many of the most capable candidates are excluded for reasons unrelated to performance. A resume may lack the precise phrasing an algorithm favors. A career break for caregiving might trigger a red flag. Skills described in different vocabulary may appear irrelevant to a machine.
The philosopher Byung-Chul Han, in The Burnout Society, describes how modern systems reduce complexity to calculable units, often flattening richness into sameness.² Hiring technologies replicate this reduction. They are programmed to reward conformity, not curiosity. A candidate with unconventional career paths, overlapping skills, or hybrid expertise is more likely to be treated as statistical noise.
One can think of uncut gems tumbling through a sorting device calibrated for size and clarity. Larger or oddly shaped stones (sometimes the most valuable when cut) are discarded because they don't resemble the preset image.
The Employer's Blind Spot
Research shows that companies relying on narrow hiring proxies, such as degree titles or tenure, often experience reduced innovation and adaptability. Narrow screening risks missing the breadth of talent available and can negatively affect business outcomes, but major studies do not attach a precise percentage to this effect.³
Insurance and financial firms, in particular, depend on foresight. Yet when systems favor yesterday's archetype, tomorrow's clients are poorly served. Homogeneity feels safe, but it is fragile. The most dangerous risk is not a bad hire, but the absence of someone who might have changed the trajectory.
The positive thread emerges here: recalibrated tools can broaden access and create a more representative workforce. McKinsey research demonstrates that diverse teams show 35 percent higher performance and 22 percent lower turnover.⁴ Employers do not just gain equity, they gain strategy.
The positive thread emerges here: recalibrated tools can broaden access and create a more representative workforce. McKinsey research shows that companies in the top quartile for gender or ethnic diversity are 35 percent more likely to have financial returns above their industry medians and up to 22 percent less likely to experience high employee turnover. These figures demonstrate the business value of a diverse workforce, where employers do not just gain equity, they gain strategy.
The Employee's Barrier
For workers, the issue is not a lack of ability, but of translation. Their skills do not align neatly with keyword maps or rigid formats. They may be multi-disciplinary, self-taught, or seasoned in roles that do not carry standardized titles. The system dismisses what it cannot immediately categorize.
Behavioral science calls this a misalignment between signal and perception.⁵ The candidate sends a signal of competence, but the system perceives it as irrelevant because it does not match the expected frame. What is overlooked is not capability, but interpretation.
Here, again, is opportunity. When filters are adjusted to evaluate broader competencies such as problem-solving, adaptability, and pattern recognition, employees gain a fairer path to consideration. Employers gain access to talent once hidden in plain sight.
Beyond the ATS
While applicant tracking systems dominate the conversation, they are only one layer. Other filters shape outcomes as well:
• Online assessments may filter out neurodiverse candidates whose cognitive approaches differ from test design, despite equal or superior performance potential.⁶
• Background-check algorithms may over flag individuals for minor discrepancies, particularly affecting candidates from different cultural or educational backgrounds.
• Keyword-driven resume screeners may exclude non-native English speakers despite fluency and relevant skills.
• Personality tests can skew outcomes by rewarding extroversion over reflection, missing introverted talents crucial for analytical roles.
These tools do not only assess, they construct the workforce by what they omit. Omission is not random—it tends to replicate historic inequities. Positive reframing is possible, however: when reassessed and fine-tuned, these same systems can reduce human bias, standardize fairness, and uncover talent across geographies.
The Inescapable Momentum
It is tempting to argue for a return to human judgment, but the momentum has swung too far. High-volume hiring demands scale. Algorithms process thousands in minutes, something no recruiter could replicate. Efficiency will not be abandoned, it will be expanded.
The question is not whether these systems remain, but how they evolve. Employers who resist adaptation risk reputational damage and narrowed pipelines. Employees who hope for purely human review will be disappointed. The path forward is to make the tools more human in design, not to remove them.
Continuous Calibration
The most viable solution is constant recalibration. Algorithms cannot remain static while the world shifts. Job roles evolve, skillsets change, markets fluctuate. Filtering systems must be audited regularly, tested for bias, and updated to ensure relevance.⁷
Employers can recalibrate in practical ways:
• Conduct quarterly audits of screening criteria for unintended bias patterns.
• Update keyword and competency libraries to match evolving roles and industry language.
• Implement secondary human review protocols for candidates marked as "near misses."
• Test algorithms on diverse demographic profiles to ensure fairness across all groups.
This is not a burden but a benefit. Employers who fine-tune their systems gain agility, matching evolving needs with evolving candidates. Employees gain fairness, knowing the evaluation adapts with context. The process becomes less about exclusion and more about alignment.
Agentic AI: A Possible Horizon
One promising frontier is agentic artificial intelligence—systems that do not simply filter by fixed rules but act with contextual awareness. Instead of scanning for keywords, such tools could interpret transferable skills, understand career pivots, and evaluate adaptability in ways closer to human reasoning.⁸
Imagine a system that not only parses resumes but cross-references them against evolving job requirements, learning in real time which qualities correlate with long-term success. Unlike rigid ATS, agentic AI could adjust to nuance, explain its choices, and refine itself continuously. It is not flawless, but it represents progress toward aligning scale with fairness.
For employers, this means uncovering talent previously invisible. For employees, it offers a chance to be considered as whole professionals rather than data fragments.
A Humble Word for Recruiters
At this point, it may sound as though machines hold all the cards, but recruiters still matter. Sometimes our role is not glamorous; it is simply noticing the person the system rejected and saying, "Wait, look again." It is not that we can override every decision, but occasionally we see potential that does not fit the filter.
Think of it as a kind of professional mischief. We present candidates who would have been discarded, not because we are heroic, but because we know that brilliance does not always match the preset mold. Recruiters may not replace algorithms, but we can occasionally outwit them, and that is often enough.
Brilliance Revealed
The final thought returns to the image of uncut gems. Many still fall through today's systems, discarded as dull because they do not fit the preset lens. But employers willing to adjust that lens, to recalibrate and refine, discover brilliance that was always there.
The machine that once erased can also reveal. For employees, that means a fairer chance to be seen. For employers, it means access to talent both diverse and capable. The stones were never lost—they were only waiting for the right light.
Sources
¹ Cappelli, P. "Your Approach to Hiring Is All Wrong." Harvard Business Review. (2019).
² Han, Byung-Chul. The Burnout Society. Stanford University Press. (2015).
³ Fuller, J., Raman, M., Sage-Gavin, E., Hines, K. "Hidden Workers: Untapped Talent." Harvard Business School. (2021).
⁴ Hunt, V., Layton, D., Prince, S. "Delivering Through Diversity." McKinsey & Company. (2018).
⁵ Kahneman, D., Treisman, A. "Changing Views of Attention and Automaticity." Varieties of Attention. Academic Press. (1984).
⁶ Austin, R., Pisano, G. "Neurodiversity as a Competitive Advantage." Harvard Business Review. (2017).
⁷ Equal Employment Opportunity Commission. "Technical Assistance Document on the Employment Provisions of the Americans with Disabilities Act." (2023).
⁸ Raghavan, M., Barocas, S., Kleinberg, J., Levy, K. "Mitigating Bias in Algorithmic Hiring." Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency. (2020).
Comments