Say you've just spent two hours on a twelve-minute interview.
It was one of those recorded interviews, the kind where a question appears on your screen, you get thirty seconds to think and ninety seconds to answer, and nobody is on the other end. You re-recorded your answer about leadership style three times because you kept hearing yourself say "um." Then you closed your laptop and realized you weren't sure you wanted the job anymore.
If you're wondering whether that reaction is unreasonable, it isn't.
And for once, I can point to something sturdier than my own twenty-five years of sitting with people in this exact moment of the job search.
It isn't you. The format is doing this to almost everyone.
A group of economists ran the experiment I have wanted someone to run for years. In A Brave New World of Hiring, researchers Mallory Avery, Edwin Ip, Andreas Leibbrandt and Joseph Vecci randomized more than 3,000 real job applicants into different interview formats: recorded audio interviews, recorded video interviews, live online interviews, and a control group. The paper was published in 2026 through the University of Exeter's economics department.
The headline finding is blunt. Asynchronous interviews (the recorded, nobody-home kind) caused an over 50% decrease in application continuation. People simply stopped going forward.
Three details in that finding matter more than the number itself.
- It wasn't only the marginal candidates dropping out. The decline held including among the most qualified applicants. The format wasn't filtering out people who couldn't cut it. It was losing people who could.
- The drop was largest for women. If you are a woman who has felt disproportionately drained by these, that is a measured effect, not a confidence problem.
- The reason was fairness, not technophobia. A companion study found the deterrence was driven by applicants' perceptions about how competitive and how fair the process was. People weren't afraid of the camera. They read the format as a signal about the company, and they were not wrong to read it as one.
So if a recording has ever made you want the job less, you were living the study's central finding.
Why companies keep using them anyway
Not because they don't care. Because they are drowning.
Hisayuki Idekoba, the CEO of Indeed, described the current hiring process to Fortune in September as a "vicious cycle," one where a company can have a thousand qualified applicants and still fail to hire the right person. The volume is the problem, and the volume is partly machine-generated: Resume Genius's 2026 hiring trends reporting found 90% of employers saw a rise in low-effort or spammy applications, much of it from auto-apply tools.
Recorded interviews are what a hiring team reaches for when it has four hundred applicants and two recruiters. It's triage. It just happens to be triage that your nervous system experiences as being evaluated by furniture.
And candidates are past patience with it. In a Monster survey of U.S. job seekers, six in ten said that not knowing whether a human ever looked at their resume was the most exasperating part of applying, and 54% said they would support heavily regulating or banning applicant tracking systems outright. That is not mild irritation. That is a constituency.
The part of the research that genuinely surprised me
I expected this study to confirm what everyone on Reddit already believes: that the robot is rigged against you. It didn't.
The same experiment had both professional recruiters and a commercial AI assessment tool (one used by most of the Fortune 100) score the interviews. The AI scored women and underrepresented racial minorities higher than the human evaluators did. It also predicted applicants' actual later employment success substantially better than the human recruiters did. And it was the humans whose assessments showed multiple cognitive biases.
Sit with that for a second, because it reframes the whole thing. The damage of the one-way interview isn't that a machine judges you unfairly. The damage is that the format drives you out of the running before any judging happens at all.
Which means the practical takeaway is not "refuse these because the algorithm hates you." It's narrower and more useful: the only real risk of a one-way interview is that it makes you quit.
How to get through one without doing four takes
The goal is to spend less of yourself on the format. These are the same mechanics I walk clients through for AI-led interviews generally, which I covered in your next interviewer might not be human.
- Record once, not four times. The research is clear that the people lost to this format include the strong candidates. Your fourth take is not meaningfully better than your first. It's just later, and it costs you the willingness to continue.
- Say "um." Leave it in. Nothing in the evidence suggests polish is what's being measured, and the human reviewers are the ones carrying the biases anyway.
- Treat it as an audition, not a conversation. You are not building rapport, so stop trying. Answer the question, give one specific example with a number in it, stop talking. Ninety seconds is enough for exactly one story.
- Set up once and reuse it. Same chair, same lamp, same blank-ish wall, laptop raised to eye level. Removing the staging decision removes most of the dread the second time.
- Decide your time limit before you start. Give yourself a budget (thirty minutes, say) and when it's spent, submit what you have. This is the same boundary-setting that protects you across a long search, which I wrote about in how to protect your mental health during a long job search.
You are allowed to ask for a person instead
Job seekers are increasingly declining the recording and asking for a short call instead, and reporting that it sometimes works. I want to be honest about this: it is a real tactic with a real cost. Some companies will move you to a call and you'll have learned something good about them. Some will quietly drop you.
My view, after a quarter century of this: it's a reasonable ask at a senior level, for a role you're genuinely strong for, where you have some leverage. It's a poor bet for a competitive entry-level posting where you are one of six hundred. Ask warmly, offer flexibility, and don't make it a stand on principle. Make it a scheduling preference.
And if you suspect the posting isn't real in the first place, that's a different problem with a different fix. I covered how to tell in which job postings are actually real.
What this is really about
The one-way interview is the most honest artifact of this hiring market: both sides are overwhelmed, and the result is a process that systematically loses the people it most wants. You can't fix that from your side of the laptop. You can refuse to let it be the thing that ends your search.
If the recorded interviews are wearing you down, that's usually a sign the strategy needs attention rather than the technique: too many applications into too many unlikely roles, each one extracting ninety seconds of performance. That's the work I do in career counseling, and it usually starts with applying to fewer, better-matched roles with a resume ranked against the job postings you choose and the software that screens them. If you'd like to talk it through, you're welcome to book a complimentary discovery call.
You are not bad at interviews. You are being asked to interview badly.