Why your CV is lying to you

Last week, I asked a simple question on LinkedIn: How do you know that the right candidate gets invited to interview?

The answer most people give, consciously or not, is: the one with the best CV. The most experience.

And I get it. When 100 applications land in your inbox, the CV feels like the only tool you have.

But here's the problem: research shows it's the wrong tool entirely.

The 1% problem

A landmark 2022 study by Sackett et al. found that the number of years on a CV explains less than 1% of how well someone actually performs in a role.

Let that sink in. The single thing most recruiters use as their primary filter accounts for almost nothing when it comes to predicting real-world job performance.

Why? Three reasons:

Years ≠ skill. Doing something for ten years just means you were there. It doesn't mean you were good at it, or that you were paying attention.

A CV is a sales pitch. It's full of spin, selective framing, and post-hoc storytelling. People don't write CVs to be accurate. They write them to impress.

It creates bias. We are drawn to recognizable company names and prestigious universities. None of them reliably predicts competence.

But doesn't experience matter?

This was the most common pushback I got, and it's a fair one.

"Would you rather have a newly qualified surgeon or one with ten years of experience?"

Yes. But that's exactly the point. What you actually want isn't the years. You want the competencies and behaviours that come with those years: the diagnostic accuracy, the calm under pressure, the clinical judgment.

The problem is that a CV can't tell you whether those competencies are actually there. Someone can have ten years in a role at a chaotic company where best practices were ignored, and still have the same CV as someone who spent two years doing everything right and learning fast.

If you evaluate competencies directly, which is what the research recommends, then you find the person who can actually do the job. Not the person who spent the most time in adjacent proximity to it.

What about AI bias?

The second big concern raised in the comments: doesn't AI risk making bias worse, not better?

It's the right question to ask of any hiring tool.

Traditional CV screening is already full of unconscious bias toward names, schools, and employers. What we've built evaluates how a candidate reasons and responds to competency-based questions. The same questions, to every applicant, without the system knowing whether someone is called Mohammed or Magnus.

Can bias creep in through poorly designed questions or skewed training data? Yes. That's why we've built the system to be transparent and auditable, and why we continuously review it. But compared to a process where you've already formed an opinion before finishing someone's CV, it's a meaningful step forward.

The real problem no one talks about

Several people in the comments said something that resonated with me deeply:

Every recruiter wishes they had time to talk to every candidate properly: to give real feedback to every applicant, and still keep the hiring manager happy.

That equation almost never works out. Volume and time are always in conflict.

That's the gap we're trying to close. Sasha, our AI interviewer, talks to every single applicant. Same questions, same standard, no shortcuts. It evaluates how they reason, not what their CV looks like. And it delivers a clear, comparable summary to the hiring manager.

You stop guessing. You meet the right people.

Is the CV dead?

Not necessarily, and we're not trying to kill it. CVs can still provide useful context, and for some roles they remain relevant. Sasha works both as a CV-free process and alongside traditional applications. What matters is that competence, not formatting, determines who gets invited to interview.

But the direction of travel is clear. The way we've hired for the last 50 years is under pressure, and that's uncomfortable. Change usually is.

The question is whether we use that discomfort as a reason to hold on, or as a reason to build something better.