Most video platforms show you your own face for the entire call. It’s the default nobody questions, because it’s always been the default.
We think it’s worth questioning.
Jeremy Bailenson’s lab at Stanford has spent years studying why video calls are more tiring than phone calls or in-person meetings. One of the four mechanisms his 2021 theory identified is mirror anxiety: the effect of watching a live feed of your own face for hours, something with no equivalent in a physical meeting room. Nobody sits across from a mirror during an in-person conversation. On video, that’s the default setup, meeting after meeting, all day.
A 2023 study from Bailenson’s group, surveying 9,787 daily video-conferencing users, tested this directly. Mirror anxiety and three related mechanisms (feeling watched by a grid of faces, feeling physically trapped, and the added effort of producing and reading nonverbal cues) together explained 34% of the variance in reported fatigue. Mirror anxiety was one of the two strongest drivers.
The gender gap in the data makes the mechanism harder to treat as a footnote. Women in the sample reported significantly more fatigue than men, even after controlling for meeting length and frequency. When the researchers tested why, mirror anxiety was the single strongest mediator of that gap, ahead of feeling physically trapped, hyper-gaze, or the effort of producing visible nonverbal cues. Women in the sample also used significantly more first-person singular language when describing their video call experience, a linguistic signature of self-focused attention that independently predicted higher fatigue scores. Two separate measurements, one mechanism.
Bailenson’s own recommendation, stated in his original piece and repeated by the Stanford VHIL team since, is simple: hide the self-view once you know you’re framed correctly. Most platforms bury that as an option you have to go find. Given what the research shows, we think it belongs the other way around: visible only if someone chooses to turn it on, not something everyone has to opt out of.
This is a small decision next to the harder problems of packet loss and round-trip latency. But it’s the same underlying discipline: look at what the research says actually costs people, and design against that instead of against habit.