We’re Not Hiring Employees. We’re Casting Performers
Authored by: Abhishek Shah
Last year, I hired someone who interviewed brilliantly. Thoughtful answers, sharp questions, a polished presence that made the entire panel confident. Six months later, they were gone, not because they had misrepresented themselves, but because the skills that made them great at interviewing had almost no relationship to what the role actually required.
That experience forced me to confront an uncomfortable truth: most companies have built a hiring process that optimizes for performance in a 45-minute conversation, then measures themselves on a completely different outcome, whether the person can actually do the job.
These two things are not the same. And we have been confusing them for decades.
The Interview Industrial Complex
The professional interview coaching industry is now a multi-billion-dollar market. Candidates rehearse stories using the STAR method. They memorize behavioral answers to questions they know are coming. They study the company, the team, and the interviewer’s LinkedIn posts. They prepare for the audition.
This is rational behavior. If a 45-minute conversation determines your career trajectory, of course, you optimize for it.
But what are companies actually learning from this process? They are learning how well someone is prepared. How articulate they are under pressure. How well they mirror the interviewer’s communication style.
Those skills matter in some roles. In most roles, they are not doing the job.
The Data Is Uncomfortable
Research on interview validity has been consistent for decades. Unstructured interviews, the kind most organizations still use, have a validity coefficient of around 0.20 when predicting actual job performance. That is barely better than reviewing someone’s photograph.
Even structured behavioral interviews, done well, top out around 0.51.
For comparison: work sample tests and job-specific skills assessments consistently score between 0.45 and 0.54. Cognitive ability assessments sit around 0.51. The difference is that interviews feel like judgment. Assessments feel like tests.
We trust our gut in the room. The data says we probably should not.
AI Made This Worse, Not Better
Generative AI changed the calculus completely. Candidates now have access to tools that write polished cover letters, optimize resumes for any job description, and generate STAR-format behavioral answers within seconds.
In 2025 and into 2026, the gap between how someone presents in the hiring process and what they can actually do has widened faster than at any point in hiring history.
This is not a candidate integrity problem. It is a measurement problem. If you measure presentation, expect the presentation to improve. Candidates are adapting to the system companies have built. The question is whether that system is measuring the right thing.
What We Are Actually Trying to Predict
Every hiring decision is a prediction. You are predicting that this person, in this role, with this team, will deliver results.
The best predictions come from evidence that closely mirrors the actual work. A candidate who can solve the kind of problems they will face on day one has given you a real signal. A candidate who can articulate a past situation where they demonstrated resilience has given you a story that may or may not generalize.
The further the hiring process is from the actual job, the weaker the prediction.
This is why work samples outperform interviews. This is why skills assessments add a signal that resumes cannot. Not because they are harder, but because they are closer to the truth of what you are trying to determine.
The Cost We Are Not Calculating
A bad hire at the manager level costs, on average, between one and two times annual salary when you factor in lost productivity, team disruption, and the cost of rehiring. Most organizations accept this as the cost of doing business.
It does not have to be.
Companies that introduce structured skills validation early in hiring – before the interview stage, consistently report higher quality-of-hire scores, lower early attrition, and faster time-to-productivity for new hires. Not because they have found a perfect hiring tool, but because they have moved the decision closer to evidence and further from performance.
What This Requires
The shift is not primarily a technology decision. It is a cultural one.
It requires accepting that a compelling interview is not proof of job readiness. It requires building processes where skills are assessed through tasks that resemble actual work, not just conversations about past work. It requires training hiring managers to weigh assessment data alongside interview impressions, rather than treating the interview as the authoritative source of truth.
Most importantly, it requires intellectual honesty about what interviews actually measure and what they do not.
The candidates who interview brilliantly are not the problem. The process that rewards brilliant interviewing over proven capability is.
The Next Hiring Standard
The companies that build stronger teams over the next five years are not the ones that interview better. They are the ones that measure better. They will shift from asking “did they present well?” to “can they do the work?” and they will build processes designed to answer the second question, not the first.
The audition served us for a long time. It is no longer fit for purpose.
It is time to stop casting performers and start hiring people.
Author Bio: Abhishek Shah is Founder and CEO of Testlify, a skills-based hiring platform helping organizations assess candidates on actual job-relevant capabilities. He writes on hiring strategy, HR technology, and the future of talent acquisition.
