Virtual Interview Mistakes That Sink Candidates (And How to Fix Them)
A 2024 study in Scientific Reports (Hiroshima University, 38 evaluators, 12 candidate videos) found that candidates who looked at the interviewer’s face on their screen scored lower on every single evaluation dimension than candidates who said nothing at all. Just voice-only. No video. The screen-gazers ranked below silence on intimacy, decisiveness, and overall hireability. That’s a striking result, and it comes down to one mistake: looking at the wrong spot on your monitor.
Virtual interviews aren’t going away. Indeed’s hiring trends research found that 93% of employers who adopted video interviews during the pandemic plan to keep using them. That was a few years ago. The number has held. So if you’re interviewing for any technical role right now, in June 2026, you will almost certainly do it over Zoom or Teams at least once before anyone meets you in person.
The good news is that most of the mistakes candidates make are fixable in an afternoon. Here’s what actually goes wrong.
The camera gaze problem (and it’s worse than you think)
This is the one that costs people without them realizing it. You’re looking at your interviewer’s face on screen. It feels polite. It’s what you’d do in person. But the camera is positioned above that face, so from the interviewer’s perspective you appear to be staring slightly downward, never meeting their gaze.
The Hiroshima University study showed the skewed-gaze condition scored worse than voice-only across six criteria. Not slightly worse. The camera-directed candidates scored higher on every single dimension the researchers measured. If you want to simulate eye contact in a video interview, you have to look at the camera, not at the person. It feels unnatural for about the first five minutes of practice. Then it clicks.
Practical fix: put a sticky note or a small arrow directly above or beside your camera lens. Point it at the lens. During answers, look at the arrow. Glance down briefly at your screen to read expressions, then return to the camera.
Technical setup issues that candidates still get wrong in 2026
The camera angle matters more than most guides admit. A camera positioned below eye level makes you look like you’re looming over the interviewer. Not intimidating in a good way. Just awkward. Stack a few books, put your laptop on them, and get the camera lens roughly at or just above your eyebrows.
Light comes from the front, not behind you. A window behind you turns you into a silhouette. Recruiters can’t read expressions on a silhouette. A cheap ring light or a desk lamp pointing at your face from the front fixes this in under two minutes.
On internet connection: if your home WiFi has any instability, use an ethernet cable for the interview. If you can’t, know exactly how to switch to your phone’s hotspot before the call starts, not while the recruiter is watching you troubleshoot. Have the backup plan ready.
And test everything the night before, not 10 minutes before. Test your microphone. Test your camera. Open the Zoom link in a browser tab. Some companies use meeting links that require downloading a new version of an app – that four-minute update process looks very different when you’re doing it in front of a hiring manager.
Something we’ve noticed in LastRound AI mock sessions
Candidates who practice video interviews with the mock interview tool before their real screen consistently report that running a recorded practice session forces them to confront camera placement and background issues they’d completely missed on their own. Watching yourself back is uncomfortable. It’s also the fastest way to fix problems you can’t see in the moment.
Background and noise: what the recruiter actually notices
A messy background doesn’t necessarily hurt you. A distracting background does. There’s a difference. A bookshelf is fine. A pile of laundry visible over your shoulder is going to get noticed, probably remembered. A virtual background that clips the edge of your head every time you move is worse than a messy room.
For backgrounds, simple and stable beats clever. A neutral wall. A bookshelf. A closed door. If you’re using a virtual background, test it with your actual camera and lighting before the interview. The greenscreen effect degrades fast when the lighting is uneven.
For noise: if you live somewhere with unpredictable ambient noise (a street, construction, a building with thin walls), use noise suppression software like Krisp or NVIDIA RTX Voice. Both run at the system level and suppress background noise before your audio hits Zoom or Teams. This isn’t overkill. The Zoom waiting room music you hear before a call starts is not the same thing the interviewer hears on their end when their filters interact with yours.
Note dependency and what it signals
Using notes during a virtual interview isn’t cheating. But how you use them matters a lot. If a recruiter can see your eyes tracking to a script below the frame every few seconds, it reads as unprepared. Like you needed to write down the answer because you didn’t actually know it.
I’d argue this is actually a reasonable interpretation from the recruiter’s side. If you need to read a written answer to “tell me about yourself,” that’s a signal about how prepared you are for trickier questions.
The workable approach: put a few keywords on a sticky note at the top of your monitor, right below the camera. “STAR structure. 3 results. Closing question.” Not sentences. Triggers. You shouldn’t need to read, just glance.
Energy and engagement on video
Video compresses your affect. You read as calmer, flatter, and more reserved on screen than you feel in real life. This isn’t a personality problem. It’s physics – bandwidth constraints, video codec compression, screen rendering. Your expressions lose subtlety in the compression.
The adjustment isn’t to perform excitement you don’t feel. It’s to be slightly more deliberate with your expressions and reactions. Nod when the interviewer makes a point. Smile when something genuinely strikes you as good. Lean slightly forward during answers rather than sitting at maximum distance from the camera. These small signals that feel normal in a room land as engaged and interested on video. Without them, you read as checked out.
The same goes for the closing question. Ending with no question is a mistake regardless of format. In a virtual interview it’s especially visible because the call ends abruptly – there’s no lobby conversation, no elevator ride, no natural wind-down. A good question buys you three more minutes of conversation and lets you close on something other than “ok, thanks, bye.” Ask something specific about the team or role, not something you could have Googled.
The one thing most prep guides skip
Most guides tell you to do a technical test. Fewer tell you to practice the format itself. A video interview is a different performance context than answering questions in your head or typing them out. Your cadence changes. Your eye contact changes. The pacing of questions and answers feels different when you’re staring at a camera and a tiny grid of thumbnails.
If you haven’t done a recorded video practice session before your real interview, you’re going in cold. You don’t know how you look on camera, how you sound, or whether your setup reads as professional. Doing one recorded session – even talking through a few answers to your most likely questions – surfaces problems you can actually fix before they matter.
For behavioral interviews specifically, practicing with the STAR method across different question types over video is the fastest way to stop sounding scripted. And reviewing how AI-assisted screening interviews work in 2026 is worth 20 minutes of your time before any first-round screen, since more companies are now using them as an initial filter.
The technical mistakes are fixable in an afternoon. The preparation gap takes a little longer. Start with the camera.
Written by
Mahesh
Writes about AI interview tooling and candidate-side interview strategy.
