Why voice, not video
People tell the truth when no one is watching.
A camera doesn't add signal. It adds pressure. Decades of research show people carry more anxiety and cognitive load on video, and open up more freely in voice. One study even found listeners read emotion more accurately from voice alone than from video. That is the method Tadeus is built on.
A lens you have to perform for.
On video, people watch themselves being watched. Self-view keeps a small mirror in the corner of the screen, and the research on mirror anxiety shows that constant self-evaluation pulls attention away from the question and onto appearance. Add gaze awareness and there is a face to manage at every moment. Take the camera away and that whole layer of performance goes with it.
Video spends focus the answer needs.
A video call asks the brain to track tone, posture, and dozens of small non-verbal cues at the same time as the words. Neuroscientists studying video-call fatigue describe this as a real cognitive tax, which is why a day of calls leaves people drained. Voice removes the visual channel, so attention lands on one place: what is actually being said.
Voice can carry more, not less.
The intuition is that seeing a face tells you more. A Yale study by Michael Kraus found the opposite for emotion: listeners read how someone felt more accurately from voice alone than from voice paired with video. The picture competes with the signal. When people feel safely unwatched, they also disclose more freely, so the record you get back is both more candid and easier to read.
| Voice only | Video | |
|---|---|---|
| Social pressure | Low. No camera, no appearance to manage, no eye contact to hold | High. Gaze awareness and self-view trigger constant self-evaluation |
| Cognitive load | Low. One stream of attention, focused entirely on what's said | High. The brain scans for missing non-verbal cues (Zoom fatigue) |
| Candour | High. People feel safely unwatched, so they disclose more freely | Lower. People mask discomfort to look composed on camera |
Sources: Yale (Kraus, voice-only emotion accuracy) · mirror anxiety & self-view research · neuroscience of video-call fatigue.
What this means for Tadeus
We chose the method that gets a straighter answer.
Tadeus runs every conversation in voice only. There is no camera, no self-view, and no second human sitting in. The employee speaks to an agent that listens, follows up, and moves on, which is closer to thinking out loud than to being interviewed.
It also keeps us clear of the line the EU AI Act draws around emotion inference in the workplace. We work from the transcript, not from a face. The record reflects what was said and understood, which is the part you can act on and defend.
Hear the difference yourself. Run one conversation.
Spin up a free Tadeus account and put a real question to your own people. The method is the easiest part to feel.
Or book a demo for your team