Who Actually Finishes an Async Video Interview

When a hiring process loses people, the usual question is how many. The more useful question is which ones.

Async video screening is a good place to ask it. The idea is simple. Instead of booking a 15-minute intro call, you send the candidate a link, they record answers to a few preset questions whenever it suits them, and a recruiter watches later. A lot of companies adopted this between 2022 and 2025, and for an understandable reason. Recruiters were spending their days coordinating calendars for calls that often could have been a form. Async video promised to take the calendar out of the first step.

It did that. It also turned the first step into something many candidates don't finish. The completion rate usually quoted for async video screens is somewhere between 40 and 60 percent. So even in a good case, about four in ten people who start one never submit it, and the figure is said to be worse in competitive fields like tech, sales and logistics.

Most teams that use async video know about the drop-off. Fewer have thought about who is dropping off, and that's the part that matters.

Look at the request from the candidate's side. They've just applied, which took some motivation. Now they're asked to find somewhere quiet with a camera and talk to nobody, knowing they'll be judged on how they come across on video as well as on what they say. There's no one to ask a question of, and no sense of when or whether anyone will watch it. For someone who badly needs a job, that's an annoyance they'll push through. For someone who already has a job and is only half looking, it's a natural place to stop.

So the people most likely to complete the screen are the ones with the fewest alternatives, who have more time to put into any one application. The people least likely to complete it are the ones several employers are chasing. They'll record a video for the company they're most excited about and quietly skip the rest. Unless you're sure you're that company, you should assume you're in the rest.

That's selection working backwards. A screening step is supposed to filter toward the candidates you want. This one filters partly on spare time and on a willingness to perform for a lens, and neither of those predicts how well someone will do most jobs.

It also filters on things that aren't evenly spread across candidates. Some people have a quiet home office and a good webcam, and some have practiced talking into a camera. Others are recording in a shared apartment on a phone propped against a mug. If the reviewer is swayed at all by lighting and eye contact, and it would be surprising if they weren't, then async video hasn't taken bias out of screening. It has moved it to the very first step, in a format that's hard to calibrate because every recording is made under different conditions.

Put those together and the pool you end up evaluating is a subset of the pool you attracted, and a skewed one. By the time a recruiter watches the first clip, the process has already made its biggest decision, and nobody made it on purpose.

That doesn't make the original problem imaginary. There are far more applicants than recruiter hours, and a live call with everyone didn't scale. But I think async video went after the wrong constraint. What was expensive was the recruiter's time to have the conversation. Async video got rid of the conversation as well, and with it the thing that made candidates willing to engage in the first place.

That line of reasoning is where Asendia came from. If async video saved the recruiter's time by throwing away the conversation, the thing to try is keeping the conversation and taking the recruiter's time out of it. So instead of a link to a recording screen, the applicant gets a phone call, usually within hours of applying, nights and weekends included. It runs a structured screen against the criteria for that role, follows up when an answer is thin, and adapts to what the candidate actually says. The candidate can ask about the job and get a real answer, which a camera never gave them. The recruiter gets a qualification summary in the ATS, a ranked shortlist and the relevant parts of the conversation word for word, rather than a clip to sit through. For the longer version of why an agent like this behaves differently from a productivity tool added to a human-paced process, I wrote a post on agentic recruiting.

I'd expect a call to do better on completion, and the reason isn't mysterious. A ringing phone asks much less of someone than a recording session does. Most people pick up, or call back. The effort is social rather than logistical, and a real back-and-forth gives the candidate a reason to stay in the process that a one-way prompt can't.

If you use async video now, here's something worth doing before you look at offer acceptance or time to hire. Take your last few roles and count how many people started the video screen and never finished it. Then open a handful of their resumes. My guess is you'll find some of the candidates you most wanted, lost at a step where nobody ever saw their names.

Ready to transform your hiring strategy? Schedule a Demo with our founders today!