Video interviews: how to run them well (and what most companies get wrong)

A recruiter schedules a video call, sends a generic link, asks the same questions they would have asked on the phone, and calls it a video interview. Nothing about that process uses the format for anything the phone could not already do, and plenty about it introduces new ways to misjudge a candidate: a bad connection read as disorganization, a toddler in the background read as a lack of seriousness, a candidate who glances at notes read as dishonesty. The camera adds information. Most hiring teams have never stopped to decide which parts of it actually matter.

The habit of treating video as a minor variation on a phone screen made sense when it was a temporary accommodation during a shutdown. It stopped making sense once video became the default first conversation for the large majority of roles, permanently, with no sign of reverting. Remote hiring runs on the same infrastructure, but even companies hiring for fully in-office roles now run most of their early-stage interviews over video, because it is faster to schedule and cheaper to run than bringing someone in. That permanence is exactly why the format deserves real design instead of inherited habit.

Getting video interviews right is not about better lighting or a nicer background, though both matter more than most interviewers think. It is about understanding what the format changes structurally: what it makes easier to fake, what it makes easier to misread, and what it quietly filters for that has nothing to do with whether someone can do the job.

Why video interviews never went back to being optional

The initial justification for video interviewing was circumstantial. The lasting reason it stuck is economic. A thirty-minute video screen costs a hiring team almost nothing beyond the time itself, no travel, no conference room, no half-day blocked off, and it lets a recruiter run twice as many first-round conversations in the same week as an in-person process would allow. Once a hiring team experiences that throughput, going back to defaulting to in-person for every early stage stops looking like a quality improvement and starts looking like a self-imposed bottleneck.

The candidate side of the equation shifted just as permanently. A candidate juggling a current job, a commute, and a narrow window of free time during business hours can take a thirty-minute video call from a parked car or a closed office door in a way they simply cannot take a two-hour round trip for an in-person first round. How to reduce time to hire depends heavily on removing exactly this kind of scheduling friction, and video interviewing remains one of the most effective levers available for doing it without cutting a stage out of the process entirely.

None of that is an argument for running every stage over video forever. It is an argument for being deliberate about which stages benefit from the speed and which stages lose something real by skipping an in-person read. Most hiring teams never made that decision on purpose. They just kept doing what started as a necessity long after it stopped being one, without ever asking which stages actually need to move back.

The technical baseline candidates are silently judging you on

A candidate experiencing a laggy connection, a recruiter who cannot find the mute button, or a platform that requires downloading unfamiliar software thirty seconds before the call starts is not thinking about the hiring process in the abstract. They are forming an immediate, specific impression of what it is like to work at this company, and that impression forms before a single interview question gets asked. Employer branding is usually discussed in terms of careers pages and Glassdoor reviews, but a clumsy video interview experience is brand damage happening in real time, delivered directly to the candidate with nobody else watching.

The fix is almost entirely procedural rather than technological. Sending a calendar invite with the platform, the link, and a one-line note on what to expect well before the call, testing audio and video on the interviewer's side before the candidate joins, and having a backup plan ready when the technology fails anyway are all low-effort steps that most companies still skip. A five-minute delay caused by a broken link reads very differently to a candidate than the same five-minute delay explained upfront with an apology and a quick recovery plan already in motion.

Indeed's guidance on running effective video interviews makes a point worth taking seriously: the platform itself matters less than most hiring teams assume, as long as it is reliable and the candidate was told in advance what to expect. A familiar, mainstream platform that works consistently beats a flashier one that requires troubleshooting, every time, because the candidate's attention should be on the conversation, not on whether their camera is going to connect.

Live video vs. one-way asynchronous interviews

One-way video interviews, where a candidate records answers to preset questions on their own time with no interviewer present, solve a real scheduling problem at scale: a company fielding hundreds of applicants for a single role cannot run a live call with every one of them, and an asynchronous format lets a recruiter review dozens of responses in the time it would take to run two live calls. For high-volume, early-stage screening, that tradeoff is often worth making.

The cost is real and specific. A one-way interview removes the back-and-forth that reveals how a candidate thinks under a follow-up question, and it disproportionately rewards candidates who are comfortable performing to a camera with nobody responding, a skill that correlates weakly, if at all, with job performance for most roles. It also tends to feel impersonal to candidates, and a format that feels like submitting a form rather than having a conversation does little to build the relationship a strong candidate needs to actually accept an offer later.

The practical split that works for most hiring teams uses asynchronous video narrowly, for high-volume initial screening on a small number of clearly defined questions, and moves to live video the moment a candidate clears that first filter. A structured interview can be run in either format, but the live version is where structure actually gets tested against a real, responsive human being, and that is the stage where the format should stop optimizing purely for volume.

What breaks when you treat video like an in-person interview

Rapport builds differently over video, and interviewers who do not account for that consistently misread candidates who are simply less comfortable on camera as less confident or less qualified in person. Eye contact, in particular, works against the camera's logic: looking at the camera lens reads as eye contact to the other person, but it means not looking at their face on the screen, and most people have never been taught to resolve that tension, interviewers included.

Pacing also changes. A pause that would read as comfortable silence in a room together reads as dead air over video, where small delays and overlapping audio are common enough that candidates and interviewers alike tend to rush their answers to avoid it. An interviewer unaware of that dynamic can mistake a candidate's rushed, under-developed answer for a lack of depth, when the format itself produced the rush.

Best interview questions still apply over video, but the follow-up behavior that makes a question work well needs to be more deliberate on camera than it would be in a room. An interviewer who builds in an explicit pause after a candidate finishes answering, rather than relying on the natural cues of in-person conversation to know when to jump in, gets a noticeably more complete answer and a far more accurate read on the candidate than one who treats the video call as a faster, lower-friction version of exactly the same conversation.

Structuring a video interview so it actually predicts performance

The single highest-leverage fix for video interviewing is the same fix that improves interviewing in any format: structure. A fixed set of job-relevant questions, asked in the same order with the same follow-ups across every candidate for a given role, removes most of the variance that the video format otherwise introduces, because it stops interviewers from improvising more loosely over video than they would in person, which happens more often than most interviewers realize.

A shared scorecard, completed immediately after the call rather than from memory an hour later, matters even more over video than in person, because video interviews tend to run back to back on a calendar with almost no buffer between them, and an interviewer who waits until end of day to write notes is reconstructing five conversations from memory rather than evaluating each one on its own terms. Building in even five minutes between video interviews specifically to complete the scorecard protects the one part of the process that actually produces a defensible, comparable decision.

Recruiting metrics that actually matter applies directly here: tracking which video-screened candidates advance, get offers, and perform well once hired is the only way to know whether the video stage is actually predicting anything, as opposed to just filtering for comfort on camera. A hiring team that has never checked whether its video screen correlates with eventual performance is running a stage of the process on faith, and video is exactly the stage most likely to be quietly rewarding the wrong thing if nobody has checked.

The bias risks that are specific to video

A video call shows an interviewer things a resume and a phone screen never reveal: a candidate's home, their visible disability, their accent and how it comes through on an imperfect microphone, their apparent age, and markers of socioeconomic background that a quiet room and professional webcam setup can mask or reveal entirely by accident. None of that information is relevant to whether someone can do the job, and almost all of it is now visible by default the moment a camera turns on.

SHRM's guidance on accessibility and accommodation in hiring technology is a useful starting point here, because video interviewing creates specific accommodation obligations that a phone screen does not: candidates with hearing or visual impairments, candidates without reliable high-speed internet, and candidates uncomfortable or unable to be on camera for disability-related reasons all need a genuine alternative path through the process, not a workaround improvised on the spot by whoever happens to be scheduling that week.

The practical mitigation most hiring teams skip is simple: score the content of what a candidate says using the same structured rubric regardless of their setup, and separate that judgment explicitly from any impression formed by their background, lighting, or visible home environment. An interviewer who notices a cluttered apartment or a rougher internet connection and lets that color the evaluation of the candidate's actual answers has let an irrelevant variable into a decision that should have nothing to do with it, and naming that risk explicitly before a hiring panel starts a round of interviews measurably reduces how often it happens.

Candidate experience: what a bad video interview costs you

Candidates increasingly treat an early video interview as a preview of what working at the company will actually feel like, and a disorganized, late-starting, technically glitchy call reads as a preview of exactly the kind of disorganization they can expect once they are inside. Candidate ghosting is often framed as a candidate behavior problem, but a strong candidate who had a frustrating video interview experience and simply stops responding to the next outreach attempt is not being flaky. They are responding rationally to a weak signal about what the company is actually like to work for.

LinkedIn Talent Solutions research on candidate experience consistently finds that the early stages of a hiring process shape a candidate's overall impression of a company disproportionately, because they happen before the candidate has any other data to go on, and a bad early experience is difficult for a strong later stage to fully undo. A company that runs a smooth, well-prepared video screen and then stumbles in a later round recovers more easily than a company that fumbles the first conversation a candidate ever has with it.

The fix costs almost nothing beyond attention: starting on time, having the interviewer's camera and audio tested before the candidate joins, giving the candidate a clear sense of what to expect and how long the call will run, and following up promptly afterward regardless of the outcome. None of that requires new technology or budget. It requires treating the video interview as a real part of the candidate's experience of the company rather than a logistics step that happens to involve a camera.

Measuring whether your video interview process is working

Most companies running video interviews at scale cannot answer a basic question: does performance on the video screen actually correlate with performance in later interview stages, or with performance on the job once hired. Without that check, a video interview stage that has quietly drifted into rewarding camera presence over substance will keep running exactly as it is, because nobody is looking at the data that would reveal the problem.

The useful audit is straightforward. Pull a sample of candidates who advanced past the video stage and compare their video scorecard to how they actually performed in later rounds and, for those hired, in the job itself. A consistent gap, candidates who scored well on video but underperformed later, or candidates who scored poorly on video but turned out strong, is a direct signal that the video stage's rubric needs to change, not that the people running it need to try harder with the same rubric.

McKinsey's research on hiring process effectiveness consistently finds that the organizations getting the most value from any interview stage are the ones treating it as a measured, adjustable tool rather than a fixed ritual nobody questions. A video interview process that has run unchanged for three years, with nobody checking whether it still predicts anything, is not a mature process. It is an unexamined habit wearing the structure of one.