Video Presence Research Brief
People only feel eye contact on video when the camera is within about 5° of the other person's eyes on your screen. A webcam sitting on the top edge of a monitor is usually 10° to 16° away.
That gap is the reason you can look attentive and still come across as distracted. This brief is two pages on what the research actually shows, with every source cited.
Download the Video Presence Research Brief (PDF)
The geometry
In 2002, Stanford researchers measured how far a camera can sit from the eyes on screen before the person on the other end stops feeling eye contact. The sensitivity turned out to be lopsided:
- Above, left or right of the lens: about 1°. Almost no margin.
- Below the lens: about 10°. Viewers are far more forgiving.
- Their design rule: keep the camera within 5° of the eyes on screen, which is roughly 3 inches at a normal desk distance.
A 2025 study refined the target further: perceived eye contact is strongest when the camera sits about 2° below the eyes on screen, not dead-on.
Now compare that to where most webcams actually sit, with the other person's face centered on screen:
- Camera at face level: about 2°
- 14" laptop: about 10.5°
- 27" monitor: about 14°
- 32" monitor: about 16°
(Angles are PlexiCam estimates, face centered on screen.)
This is why "just look at the camera" never quite works: you can look at the lens or at the person, not both.
What it costs
Three findings from the brief:
Job interviews. In a 2024 study, candidates in a simulated online interview who looked at the interviewer's face on screen (about 26° below the lens) instead of at the camera (about 5°) scored lower on all six ratings, including hireability. 38 evaluators; every difference significant at p < 0.001.
Influence and negotiation. In a 2015 study, people in a real-money negotiation over video deferred to a partner filmed from below 62% of the time, and to a partner filmed from above just 38%. The same person was judged about 3 cm taller when filmed from below.
Telehealth. In a 2022 cross-cultural experiment, 88.4% of Japanese and 90.2% of Lebanese participants chose the physician who gazed at the camera over the one who gazed at the screen, and rated them higher on communication and interpersonal skills. (43 Japanese and 61 Lebanese participants.)
Five fixes, in order of impact
- Put the camera where their face is. Within about 3 inches of the other person's eyes on your screen, ideally just above them. This is the only fix that lets you look at the person and the lens at the same time.
- Get the lens to your own eye height. Within an inch or two. Too low and you loom over the viewer; too high and you look small and talked-down-to.
- Frame yourself on the top-third line. Eyes about a third of the way down the frame, a sliver of headroom, head and shoulders filling the shot from roughly an arm's length away.
- No camera you can move? Move the window. Shrink the call window and park it directly under the lens. It's free, and it's what Stanford's own prototype did.
- Be wary of AI gaze correction. Software that redraws your eyes points your gaze at the viewer while your head points elsewhere. People may not name it, but they notice. Worth knowing that the telehealth authors cited above take the opposite view and call for platforms to build it in. We'd rather move the camera than repaint the face.
What the research doesn't say
Eye contact isn't magic. A 2025 study found that deliberately staring into the lens to fake eye contact had little effect on interview ratings, and a 2022 teletherapy study found better eye contact didn't raise perceived empathy over 20-minute sessions. Forcing yourself to look at a lens is not the same as a camera placed where you're already looking.
The effects above are strongest in short, high-stakes first impressions: interviews, pitches and first meetings. The top-edge camera does work fine when the call window is small and tucked right under it, but few people use video that way on a large monitor.
How PlexiCam fixes the geometry
Mag2 plugs into the USB-C port on the back of your webcam and attaches magnetically to your screen, so the camera sits right over the face you're talking to. It works with most webcams that have a rear USB-C port and about 95% of modern monitors.
Pro V2 hangs any tripod-thread webcam in front of the screen.
Both keep the webcam you already own, and neither uses any desk space.
Sources
- Chen, M. "Leveraging the Asymmetric Sensitivity of Eye Contact for Videoconferencing." Proceedings of CHI 2002. graphics.stanford.edu/papers/eye_contact
- Gao et al. "Don't look at the camera: Achieving perceived eye contact in remote video communication." Journal of Vision, 2025. PMC12439499
- Shinya, Yamane, Mori & Teaman. "Off-camera gaze decreases evaluation scores in a simulated online job interview." Scientific Reports, 2024. PMC11143298
- Thomas & Pemstein. "What you see is what you get: webcam placement influences perception and social coordination." Frontiers in Psychology, 2015. PMC4365549
- Helou et al. "Physician eye contact in telemedicine video consultations: A cross-cultural experiment." International Journal of Medical Informatics, 2022. PubMed 35809542
- Jelson et al. "Investigating the Effects of Simulated Eye Contact in Video Call Interviews." CHI 2025.
- Grondin et al. "Clients in Simulated Teletherapy via Videoconference Compensate for Altered Eye Contact When Evaluating Therapist Empathy." Journal of Clinical Medicine, 2022.
Published by PlexiCam, October 2026. We make camera mounts, so we have a stake in the conclusion. That's why every claim above is sourced and the limits are stated.