AI Candidate Screening and the Limits of the Resume

A resume is an odd thing to base a hiring decision on. The candidate writes it, it's tuned to get past whoever reads it first, and it mostly tells you where someone has been rather than what they can do. That was tolerable when finding out more about every applicant was expensive. I think it's getting cheaper, which is a large part of why companies now talk about skills-based hiring.

The weaknesses of resumes are familiar. They reward pedigree over proficiency, so a well-known school or employer reads as strength whether or not the person is good at the work. They're easy to dress up. And they do a poor job of showing skills people picked up outside the usual path, say in a job with an unhelpful title or on the way through a career change. The people who get passed over are often capable ones who don't fit the usual mold.

We made a version of this argument in The Resume Is Dead. What a recruiter wants to know sits underneath the polish of a CV: whether this person can actually do the job, and how much they could grow into it.

Older screening software didn't get you much closer, because it mostly matched keywords. If the right words were on the page, the candidate passed, so the best resume writers won. Newer AI screening tools use machine learning and natural language processing to read more the way a person would. They look at the skills someone has shown and the experience behind them, and they try to judge growth potential too.

That makes skills-based hiring, which means hiring for what people can do rather than where they've been, much more practical. A good AI screen can notice when a skill learned in one setting carries over to another, which widens the pool of people you'd consider. It can look for evidence of how well someone learns and adapts, which matters more in fields that change quickly. And it can hold every candidate to the same set of requirements.

That last one is where fairness comes in. People reading resumes are inconsistent, and they carry preferences they may not know they have. A process that asks every candidate about the same things and scores them the same way removes a lot of that bias. I'd be careful not to overstate it, though. Consistency and fairness aren't the same thing, and an AI screen is only as fair as the criteria you give it. But treating every applicant the same way is a much better starting point than depending on who happened to open the file.

If most of your screening still runs through the ATS, it's worth being honest about what that system was built for. We've written about its limits in Your ATS Isn't a Sourcing Tool.

Screening is what we work on at Asendia. We make software that phones applicants and interviews them, so I obviously have a view. But the argument doesn't depend on our product. If the resume mostly measures pedigree and polish, and what you care about is whether someone can do the job, you need another way to find out. The best answer I know is a consistent look at actual skills, and AI is what makes that affordable to do for every applicant instead of only the few whose resumes stood out.

Ready to transform your hiring strategy? Schedule a Demo with our founders today!: https://asendia.ai/talk-to-founders