Why Human Judgment Still Matters in Hiring
Every few years, a new tool promises to take the human bias out of hiring and replace it with something more rigorous. The tool always underperforms on the thing that matters most: predicting whether a specific person will succeed in a specific role at a specific company. Here is why that prediction remains stubbornly human.
Founder, Majhi Group & Majhi OS
I have been doing executive search long enough to have seen several generations of technology promise to transform the assessment of human candidates. Structured interviews. Personality assessments. Predictive hiring algorithms. Skills-based evaluations. AI sourcing tools. Video interview analysis.
Each of these tools has genuine value in specific contexts. Each of them has also produced a version of the same mistake: the confusion of correlation with causation, of pattern-matching with judgment, of what can be measured with what matters.
I want to be precise about this, because I run a hiring operations company that uses technology throughout the search process, and I don't want to be read as a technophobe defending an old model. AI tools have made parts of what we do meaningfully better. What they have not changed, and I believe cannot change in the near term, is the core judgment that determines whether an executive placement succeeds.
What the judgment actually is
When I assess a candidate for a VP or C-suite role, I am not primarily evaluating their resume. The resume is a starting point, not a conclusion.
What I am evaluating is a much harder set of questions: How does this person think about a problem I put in front of them? What do they reveal about themselves when they're uncertain versus when they're confident? How do they talk about the people they've managed, with specificity and genuine insight, or with generalities that suggest they weren't paying close attention? What does the delta between their public track record and what their references say tell me about their self-awareness?
These are not inputs to a model. They are observations that I am making in real time, filtered through everything I've seen in hundreds of similar situations, and combined into a judgment that is irreducibly subjective, but not arbitrary. Subjective means it depends on a subject. In this case, the subject is an experienced professional who has been wrong, learned from being wrong, and developed a more calibrated instinct over time.
The AI tool does not have that calibration. It has pattern matching against historical data. That is useful. It is not the same thing.
The assessment that determines whether an executive placement succeeds is irreducibly subjective - but not arbitrary. AI has pattern matching. That is not the same thing.
Where technology genuinely helps
I want to be fair to the tools because I use them and they matter.
Technology has made candidate sourcing dramatically more efficient. What used to take a research team a week to compile, a qualified list of candidates in a specific functional area, in a specific geography, at a specific company stage, can now be done in hours with the right tools. That efficiency is real and it allows the human judgment to be applied to a better-curated set of candidates rather than to a noisier, less organized universe.
Technology has also improved the administrative layer of search: scheduling, communication, documentation, status tracking. These are areas where human error and inconsistency create friction, and where systematization improves the client and candidate experience without requiring any of the judgment I've described.
The right frame is: technology is raising the floor of what a search process produces, and freeing human judgment to focus on the ceiling. The sourcing, the scheduling, the first-pass screening, these can be better done with good tools. The final assessment, the candidate advocacy, the negotiation, the offer management, these require humans who know what they're doing.
The fit judgment that cannot be automated
The hardest judgment in executive search, the one where I've been most often right and most consequentially wrong, is cultural and contextual fit.
A candidate can be excellent in an absolute sense and wrong for a specific company at a specific stage. I have seen brilliant executives fail in companies where the culture required a different operating style. I have seen people who looked marginal on paper succeed enormously because the company's specific challenge was the one they were built for.
That judgment, whether this person, in this company, at this moment, with this team, is not derivable from a model. It requires understanding the company's culture from the inside, which means the kind of relationship with the CEO and leadership team that you build over years of work together, not from a brief. It requires understanding the candidate as a person, not as a profile.
When I make that judgment call and I'm right, it produces placements that the client describes as transformative. When I get it wrong, which I have, and which any honest search practitioner will acknowledge they have, the damage is significant: a failed executive hire at the VP level costs a company 12-24 months and anywhere from $500K to $1M+ when you account for the direct cost, the lost productivity, and the team disruption.
The cost of getting it wrong is why the judgment matters, and why the judgment needs to be human. Not because humans don't make errors, we do, regularly, but because the errors we make are recoverable in a way that the errors a model makes are not. When I get it wrong, I can explain why, I can learn from it, and I can do the relationship repair that a failed placement requires. When an algorithm gets it wrong, there's no relationship, no accountability, and no recovery path.
The festival of context
Here's an illustration that has stayed with me.
Several years ago, I was placing a senior finance executive at a company going through a difficult restructuring. On paper, the candidate was strong: the right functional background, the right sector experience, strong references. In the interview process, he was excellent.
What I noticed, and what I flagged to the client before the offer was extended, was a subtle pattern in how he talked about ambiguity. When I asked him about the last time a major plan changed significantly and he had to adapt, the structure of his answer was technically complete but emotionally thin. The words were right. The affect was slightly off, a quality of rehearsal rather than genuine recollection.
I recommended they do one more reference call, with a specific question about how this person handled a major plan failure. The reference, which the candidate had provided, said something between the lines that confirmed my read: "He's excellent when the direction is clear. When things are truly unclear, he needs more structure than you might expect."
That was a deal-changer for a restructuring role. The client went in a different direction.
I could not have told you, before that conversation, exactly what I was looking for. There was no rubric. It was pattern recognition accumulated from hundreds of conversations with people at that level, filtered through the specific requirements of that role and that company. The technology helped me find him. The judgment is why we didn't place him.
Majhi Group's searches are built around the judgment that tools make more efficient, not replaceable. Request an assessment to discuss a current leadership search.
See also: AI and the Future of Talent Sourcing, AI and Hiring: What Actually Changes, The Global Hiring Floor
Sources
HBR: How to Take the Bias Out of Interviews
Frequently Asked Questions
Can AI replace human judgment in executive hiring?
Not at the level that matters most. AI tools can improve the efficiency of the search process, sourcing candidates, screening resumes, scheduling interviews, analyzing patterns across large datasets. What they cannot do is make the judgment call that determines whether a specific person will succeed in a specific role at a specific company. That judgment depends on reading people in context, understanding the specific culture and stage of the hiring organization, and applying experience from hundreds of similar situations to a combination of factors that is always unique. That is human work.
What are the limitations of AI in the hiring process?
AI excels at pattern matching across large datasets: finding candidates who match historical profiles of success, screening at volume, and reducing the administrative load of the search process. It struggles with: situations that don't match historical patterns (a company at a genuinely new stage, a role that has never existed before); the qualitative assessment of culture fit and leadership presence; the judgment call on candidates who are strong on some dimensions and weak on others in ways that don't reduce to a score; and the relationship management that is often the difference between a candidate accepting and declining an offer.
How does an experienced search professional assess a candidate differently than an AI tool?
An experienced search professional is assessing things that don't appear in a resume or a structured interview: how a candidate talks about their failures, how they respond when a question reveals something unexpected about the role, whether their confidence tracks their actual track record or exceeds it, how they treat support staff during a site visit, what their references say between the lines. These signals are rich, contextual, and require accumulated pattern recognition about what people look like at different levels of genuine capability. They are not systematizable in the way that resume screening is.
Did this land? Push back? Add something I missed?
Reply to Manas →Continue Reading
Related writing
Future of WorkAI and the Future of Talent Sourcing
AI has already changed how candidates are found. What it hasn't changed — and what I believe it won't change in the foreseeable future — is how the best candidates are persuaded. The sourcing problem was never primarily about finding people. It was about reaching the right people in a way that makes them want to engage.
Future of WorkThe Global Hiring Floor
AI is not just changing who does the work. It is changing what the work is worth. As AI tools raise the output floor for every knowledge worker, the question of what human talent commands, and why, is being rewritten in real time.
Future of WorkAI and Hiring, What Actually Changes
AI is changing hiring. Most of the conversation about how is wrong. The changes that matter are not about automation replacing recruiters, they are about who wins the talent competition in a world where sourcing is commoditised.