Low-fidelity CV

Where were you when the perfect job was looking for you? Where were you when the hiring manager and the recruiter needed someone just like you? You were probably hiding behind your CV.

CVs have always been a terrible proxy for performance. Just to cite a piece of widely available research:

Years of experience and job titles, the metrics most recruiters quickly look for on a CV, have between 0% and 3% predictive validity when it comes to actual job performance. For comparison, even a loose, unstructured interview has around 20%.

So it’s not only that CVs don’t tell the full story but, more importantly, that they miss the most critical part entirely. For organizations aiming to make hiring their competitive advantage, CVs have become more of a distraction than anything else.

Great hiring needs to be based on evidence, and a CV is evidence of nothing. Even the best-tailored resume from the most honest candidate, complete with the right keywords, metrics, the XYZ format, and a dedicated section for top skills, all capable of supercharging even the most hidden unconscious biases of the recruiter, does not solve the fundamental problem. There is a massive gap between claiming a skill and demonstrating it.

Everyone knows this gap exists. Yet CVs remain the epicenter, the mandatory starting point of nearly every hiring process.

And that’s an unfortunate and fascinating paradox: we use a hiring artifact that we know is scientifically weak to make some of the most expensive decisions organizations face.

Why? I’m not sure. Maybe because it costs nothing to receive a CV in PDF format. Or because, while it has almost no predictive validity, it at least gives you basic relevance, quickly confirming that a candidate is a doctor and not a plumber. Or perhaps because it is convenient. CVs offer a standardized format that makes comparison and filtering easier, and reimagining the system feels too expensive, especially when the entire ATS and broader talent acquisition stack is already built around resumes.

On top of the historic challenges with CVs, AI is making things even trickier and resumes even less trustworthy. The modern AI-driven recruiting tech stack is stuck in a loop where AI generates resumes and applications, and AI scans and screens them, while human recruiters struggle to determine who is a real candidate anymore. Even in voice and video interviews, trust has eroded, especially in entry-level technical roles.

In this environment, Ashby noted 18 months ago that referrals and internal hiring are resurging.

Due to AI-powered high-volume one-click applications, referrals have 10x better conversion from “Application” to “Offer” compared to normal applicants. Their platform has even built specific features to support “Internal First” workflows, recognizing that an internal performance review is a “high fidelity” artifact, whereas a resume is a “low fidelity” artifact.

This is happening not necessarily because referrals and internal candidates are better, but because they act as proof of humanity, a form of social proof, which is a strong signal that something is broken.

Is hiring broken overall? I don’t think it is. Many organizations are constantly seeking better ways to identify and recruit great people. Many of them are highly sophisticated in how they hire, developing better processes, more focused interviews, more relevant assessment frameworks, and clearer role definitions. The resume is one of the obstacles they face.

The big question isn’t whether hiring is broken, but whether we’re willing to let go of resumes as the default entry point. What could replace it as the default starting point? What’s the solution? What are the organizations that make hiring a competitive advantage actually doing?

They are adopting a multi-method approach, combining logic tests, structured interviews, and work samples.

This approach, which is core to skills-based and performance-based hiring, has significantly higher predictive validity than CVs. Research consistently supports this, and organizations have increasingly adopted these methods in recent years to counterbalance the resume and determine whether candidates are who their CVs claim they are.

Just FYI, skills-based and performance-based hiring aren’t new. In fact, they have long been supported by evidence. They failed to become standard practice not because they were wrong, but because they were too complex to execute at scale. Jobs evolve constantly. What “great performance” looks like changes by company, growth stage, and team context. A hotel GM role in an established luxury resort in the Bahamas requires different skills than the same role in a new hotel in the suburbs of Cleveland. Skills frameworks require continuous maintenance and updating. It was simply too hard to keep up.

This is where I think AI can actually help. Rather than further optimizing the resume, AI can help organizations build more effective ways to recruit, identify, and hire extraordinary people. In other words, it could strengthen the signal rather than amplify the noise.

For skills assessment, AI could help organizations define and update skill requirements dynamically, map experiences to competencies more rigorously, maintain evaluation frameworks over time as roles evolve, and infer required skills from job context.

For performance evaluation, AI can help organizations stay disciplined in structured hiring and support interviewers in asking consistent, evidence-seeking questions tied to clearly defined success criteria.

What if we stopped automating bad inputs and reimagined the first step entirely? If we removed the resume, what would replace it?

Here is one AI-driven alternative:

An organization might start by clearly articulating the role, defining what strong performance looks like, and identifying the skills and experiences that matter most in that specific context. Candidates could then respond to a focused set of structured questions designed to elicit evidence of past behavior and results. From those responses, a standardized and comparable profile could be generated.

This could replace a CV, a document created by candidates, with a standardized, comparable skills experience profile based on what the hiring organization actually needs, what the job truly entails, and the competencies required for success.

Changing the artifact that anchors the process would likely change the conversation itself. It would not eliminate uncertainty or bias entirely, but it could shift attention toward evidence and away from claims and aesthetics. Better yet, it might surface capable individuals who are currently overlooked because their resumes fail to communicate the depth of their abilities.

If we are serious about improving how organizations identify and select talent, we may need to question not only how we evaluate candidates, but where we begin. Perhaps the real issue is not that hiring is broken. It is that we have been starting from the wrong place.

Thank you for reading, and don’t forget to subscribe if you want to know when the next post is out.