Social Life in the Metaverse Is Weirder Than Expected

The Avatar’s Uncomfortable Stare

You are standing in a virtual room, wearing a heavy headset. Across from you, a colleague’s avatar shifts its weight. Its eyes do not quite meet yours. You feel a flicker of something: not quite connection, not quite alienation. A strange, in-between state.

This is the core paradox of the metaverse. We assumed that moving from flat Zoom boxes to immersive 3D spaces would make social interaction better. More natural. More human. But a major 2022 study by Thorsten Hennig-Thurau and his colleagues at the University of Münster suggests something far stranger is happening. The metaverse does not simply improve social interaction. It transforms it into something fundamentally different, with its own weird logic, its own unexpected failures, and its own surprising strengths.

The researchers, publishing in the Journal of the Academy of Marketing Science, ran a series of field experiments comparing real-time multisensory social interactions (RMSIs) in virtual reality against those on traditional 2D platforms like Zoom. They wanted to know: does putting on a headset and embodying an avatar actually create more value for people interacting?

The answer, it turns out, is a deeply qualified yes. And that qualification is where the story gets interesting.

The Headset Does Not Fix Awkwardness

The first surprise came when Hennig-Thurau and his team compared how people performed on a collaborative task. In one condition, participants interacted through VR headsets. In the other, they used a standard video conferencing setup. The task required them to work together to solve a problem.

The researchers found that VR did not automatically boost performance. In fact, the initial results were contradictory. Some groups in VR performed better. Others performed worse. The authors write that their field-experimental probes “support the value-creation potential of the virtual-reality metaverse, but contradict its general superiority” (Hennig-Thurau et al., 2022).

This is not what the hype promised. The metaverse was supposed to be an upgrade. Instead, it seems to be a trade-off.

What explains the inconsistency? The researchers dug into the intermediate conditions: the psychological states that people experience during the interaction. They measured things like social presence (the feeling of being with another person), physical presence (the feeling of being in a place), and cognitive load (how hard your brain is working to process the experience).

What they found is that VR excels at creating a sense of physical presence. You really feel like you are in the virtual room. But this comes at a cost. The same technology that makes the space feel real also makes the social interaction feel more demanding. Your brain is working overtime to parse the avatar’s movements, to adjust to the slight latency, to reconcile the mismatch between what you see and what you feel.

This cognitive overhead can undermine the very thing the metaverse is supposed to improve: the quality of the social bond.

The Uncanny Valley of Social Interaction

The problem is not just technical. It is psychological. When you interact with someone through a video call, you are looking at a representation of their actual face. The mismatch between the image and the person is minimal. You know it is a screen. You adjust accordingly.

But in VR, you are looking at an avatar. And avatars are weird.

The researchers’ framework suggests that the metaverse creates a new kind of social space where the usual cues are both amplified and distorted. Eye contact becomes a technical problem. Body language becomes a design choice. The avatar’s face may smile, but the timing is slightly off. The result is a feeling that social psychologists might recognize as a cousin of the uncanny valley: not quite human, not quite machine, but something in between that triggers a subtle unease.

Hennig-Thurau and his team call this the “intermediate conditions” that mediate between the technology and the outcome. They identified three key mediators:

• ▸Social presence: The sense that you are interacting with a real person. VR can increase this, but only if the avatar is well-designed and the interaction is smooth.
• ▸Physical presence: The sense of being in a shared space. VR excels here. But too much physical presence can actually distract from social presence.
• ▸Cognitive load: The mental effort required to process the experience. VR increases this load, which can degrade performance and emotional response.

The authors found that these factors interact in complex ways. A high sense of physical presence might boost your emotional engagement, but if it also increases your cognitive load, you might actually perform worse on a task. The net effect depends on the specific combination.

The Moderation Trap: When VR Backfires

The study also uncovered something more troubling. Not everyone benefits equally from the metaverse. The researchers identified several moderating factors that can flip the effect from positive to negative.

One of the most important is the nature of the task itself. For tasks that require high levels of coordination and rapid feedback, VR can actually be worse than video. The slight latency, the imperfect tracking, the need to adjust to the avatar’s movements: all of these create friction that disrupts fluid collaboration.

Another moderator is the quality of the avatar. If your avatar looks like a cartoon character with dead eyes, the social presence effect collapses. You are not interacting with a person anymore. You are puppeteering a doll. And that feels different.

The researchers also found that individual differences matter. Some people are more susceptible to cybersickness. Some people are more sensitive to the uncanny valley. Some people simply prefer the flat, predictable world of video calls. The metaverse is not a universal upgrade. It is a specific tool that works best for specific people in specific contexts.

This is a crucial insight for anyone building metaverse products. You cannot just throw people into VR and expect magic. You have to design for the intermediate conditions. You have to manage cognitive load. You have to make avatars that do not creep people out. And you have to accept that for many use cases, Zoom might actually be better.

The Research Roadmap: What We Still Do Not Know

The authors end their paper with a research roadmap. They are honest about what they have not proven. The study is based on field experiments with specific tasks and specific populations. It is a starting point, not a conclusion.

One major open question is about long-term effects. The experiments measured immediate outcomes. But what happens after a hundred hours in VR? Does the cognitive load decrease as you habituate? Or does the weirdness compound? We do not know.

Another question is about the role of haptics and other sensory inputs. The current study focused on visual and auditory VR. But the metaverse promises touch, smell, even taste. How will those additional channels change the social dynamic? The authors explicitly call for research on “multisensory” interactions beyond just vision and sound.

There is also the question of scale. The experiments involved pairs or small groups. What happens when you have a hundred avatars in a virtual auditorium? Does social presence scale? Or does it degrade into a cacophony of cartoon figures?

And perhaps most importantly, the study does not address the ethical dimensions of avatar-mediated interaction. If your avatar can be manipulated by the platform, if your eye contact is being tracked and optimized, if your emotional responses are being measured and fed back to you, what does that do to the authenticity of the interaction? The authors flag this as a “societal challenge” that deserves attention.

The paper does not claim to have all the answers. What it does is provide a framework for asking better questions. And the first question is this: what are we actually trying to achieve when we move social interaction into the metaverse?

What This Actually Means

The research by Hennig-Thurau and his team is not a verdict on the metaverse. It is a diagnostic. It tells us where the technology works, where it fails, and why. Here is what that means in practice.

• ▸Do not assume VR is better for collaboration. For tasks that require fast, fluid coordination, video calls may actually outperform VR. The cognitive load of the headset can slow you down. Save VR for tasks where physical presence matters more than speed: creative brainstorming, spatial design, or social bonding.

• ▸Avatar design is not cosmetic. It is functional. A poorly designed avatar kills social presence. The uncanny valley is real. Invest in making avatars that look and move like actual humans, or accept that your users will feel a persistent, low-grade weirdness that undermines trust.

• ▸Manage cognitive load deliberately. The metaverse is not a passive medium. It demands attention. Design interactions that minimize unnecessary complexity. Short sessions. Clear goals. Minimal distractions. The more your users have to think about the interface, the less they can think about each other.

• ▸Measure intermediate conditions, not just outcomes. If you are running a metaverse pilot, do not just ask whether people liked it. Ask whether they felt present. Ask whether they felt overloaded. Ask whether the avatar felt like a person. These intermediate states predict long-term engagement better than any satisfaction score.

• ▸Accept that the metaverse is not for everyone. Some people get sick. Some people get creeped out. Some people just prefer the flat, reliable world of screens. That is not a failure of the technology. It is a constraint. Design for the people who benefit, and do not force the rest.

The metaverse is not a better version of the internet. It is a different version. And different means trade-offs. The researchers have given us a map of those trade-offs. Now it is up to us to decide where to go.