H

i there, I’m Devin Mancuso, and I’m a Senior Design Strategist at Google. Back in 2018, I was the UX Lead for a team at YouTube focused on looking at the emerging phenomenon of virtual influencers appearing on YouTube, or “VTubers” (Virtual YouTubers) as they affectionately referred to themselves as. I was fortunate enough to have the opportunity to interview a number of VTubers across Japan, the US, and the UK over the course of the year.


I recently gave a talk at Pause Fest 2020 in Melbourne, Australia, discussing VTubers, autonomous virtual influencers, and the emergence of new video media platforms. This article is a collection of insights from that talk. You can find the full recording at the bottom of this article.


What are VTubers?

So, what are VTubers? VTubers, or Virtual Youtubers, could be considered a subset of virtual influencers. VTubers are social media personalities who publish content as 2D or 3D characters that are almost entirely puppeteered and voiced by humans. The term originated because many of these creators began publishing content on YouTube, but these days, VTubers are active across many different social platforms.


So VTubers are virtual influencers, but are virtual influencers VTubers? Honestly, the nomenclature is challenging at the moment: virtual influencers, digital humans, CGI creators, VTubers… the edges are blurry with a lot of overlap. However, during my research, I noticed a defining characteristic of VTubers was that they were typically puppeteered or ‘driven’ by a human actor in some form of real-time capture. This is in contrast to a virtual influence who, in some cases, might only be a 3D model posed and rendered for static image content on Instagram.


So when did VTubers start appearing?

While there have been a number of creative projects on YouTube over the years featuring digital characters as online personalities, including a project from Mattel featuring Barbie as a Vlogger back in 2015, VTubers as we know them today owe a lot to the breakout success of one particular VTuber, Kizuna AI.

Kizuna AI

Kizuna AI is a Japanese VTuber with a following of more than 4 million subscribers across three YouTube channels (not to mention another 1 million subscribers on Bilibili), making her the most popular VTuber to date. She uploaded her first video to YouTube in November 2016, inspiring an entire generation of VTubers to follow. These days, Kizuna maintains an active social presence online and offline; walking around Tokyo, it is not uncommon to see Kizuna featured on billboards, posters, or merchandise across the city.


Fast forward to 2020 and the landscape is quite different now, with thousands of active VTubers on YouTube. In 2017 the VTuber community was largely comprised of Japanese creators, but today many VTubers from around the world have launched their own VTuber personas and unique character designs.


VTubers as YouTubers

When people first learn about VTubers, they often challenge the concept: “Aren’t these just cartoons or 3D characters? I’ve seen cartoons on YouTube for years!” However, while cartoons and animated characters have existed on YouTube since its inception, what’s most different and most unique about VTubers is the characters present themselves as real entities, existing in our world, on our timeline, alongside human content creators.


VTubers want to emulate the style popularized by human YouTubers over the past decade. Further, they don’t necessarily want to be perceived by their audience as any different from their human counterparts. See Lilly Singh alongside Kizuna AI here:


VTubers already create content akin to human YouTubers today: vlogging, game streams, music videos, educational content; even unboxing and product reviews. These virtual creators are also heavily involved in the YouTube community, not to be outdone by their human counterparts, they even participate in viral YouTube trends such as the Spicy Packet Noodle Eating challenge pictured below.

VTubers want to take on any viral challenge a human can, even spicy noodles!

Why do people choose to perform as a VTuber?

So, why do these content creators choose this format to share themselves with the world? Well, after chatting with a whole cast of VTuber creators across the globe, the following are some of the most common reasons they mentioned:


  • It’s Fun! These creators enjoy the creative freedom they get from VTubing and exploring genres and content formats they might not normally pursue or even be able to create as their human selves.
  • Convenience: It’s incredibly convenient. VTubers often prefer this medium because it allows them to bypass things like a home studio setup, or having to do hair and makeup before getting in front of a camera. They can literally perform from bed using these technologies.  
  • Anonymity: Many creators are turning to VTubing as an opportunity to publish online without having to reveal their human likeness to the public. This might be because a preference to not have their face on camera or because the context means they might be put in danger if their identity was known communicating their message.
  • Brand Strategy: VTubing is a way for these creators to experiment and extend their brand in a brand-safe way online. Using a virtual character allow brands to experiment with new markets and content genres without associating their human talent until they have proven the space works for their brand.


Why do audiences like to watch VTubers?

Audiences appear to enjoy watching VTuber content for reasons similar to that of traditional YouTube content: they want to be entertained, educated, informed, and inspired.


However, there is one extremely unique and novel element to VTuber content that draws in and engages audiences: a sense of play. Watching VTuber content is, in a way, more participatory than traditional human YouTube content. This is because, in most cases, the audiences also gets the opportunity to “play along”, suspending disbelief and participating in a grander narrative in which this virtual character exists in their world too.


This kind of grand narrative plays out all over our social platforms, it’s as if the characters from our favorite shows have leapt out of the screen and into our world. When the YouTube video is over, these characters then go and post on Twitter, update their Instagram stories, and interact with fans via comments.


This construct likens itself to the idea of kayfabe. The term kayfabe traces its origins back to professional wrestling, where it is described as the portrayal of staged events within the industry as "real" or "true", specifically the portrayal of competition, rivalries, and relationships between participants as being genuine and not staged.


In the case of VTubers this form of kayfabe carries with it a social contract between the VTuber and their audience; 2 key rules must be followed:


  1. For the VTuber. Stay in character. VTubers must stay in character, they must maintain the illusions and act consistently with the narrative history they have created for their persona. Most VTubers will not acknowledge that they are not real if questioned.


  1. For the fans. Play along. Fans are expected to remain emotionally faithful, suspend disbelief, and play along ‘with the gag’. Sometimes this involves defending the narrative/creator in the comments, or shepherding and providing guidance to new fans in the comments or live chat who aren’t quite sure what they’re witnessing.


When both sides fulfill their end of the contract, an emotional fidelity is achieved that is incredibly captivating for audiences. In exchange for remaining emotionally faithful, audiences get to feel something real, something they know is real because it originated from inside them: joy, jealousy, admiration, pride. This phenomenon goes some of the way to explain why fans are often unbothered by factual claims that the VTuber is “not real” or “controlled by a company”. The emotional fidelity is far greater than the factual fidelity for these fans.


Maintaining this kayfabe is no easy feat however, it is a form of emotional labor for the VTubers themselves. Audiences demand authenticity and fans expect VTubers to create a faithful emotional reaction in them that they know to be real.


Ultimately, while novelty brings audiences, “what is this thing…???”, it's the personality that keeps audiences coming back. Viewers expect the same level of authenticity from VTubers that they demand from human influencers across social media.


What’s next?

In the second half of my talk [11:20], we explore emerging technologies in the space that could lead to the development of autonomous, virtual influencers. We explore 3 interesting technology areas that are actively being worked on by companies and academics in this space: avatar diversity, real time rendering, content generation.


We also discuss [12:20] the potential for autonomous virtual beings and technologies like cloud rendering to lead to the development of new media platforms that challenge our traditional definitions of video media. We explore how this could fundamentally alter the relationship between content creators and viewers shifting from a 1-to-many relationship model to a series of 1-to-1 relationships.

Today, creators maintain a 1-to-many publishing relationship with their fans.
In the future, dynamic videos coupled with autonomous influencers might see a radical shift to a 1-to-1 relationship between creator and fans.

Lastly, we look at [12:15] how VTubers might one day use real time data from their audience to alter and personalize their performance on the fly, creating the potential for a hyper-personalization loop between creator and fan.

Dynamic performances might be underpinned by hyper personalization

For more on VTubers and virtual influencers, you can watch the full talk below, recorded at PauseFest 2020 in Melbourne, Australia.

Many thanks to the YouTube VR/AR team and to all of the VTubers across the globe who generously made time to speak with me so I could learn more about their amazing work.

Concepts Discussed
Virtual Influencers Mentioned
Kizuna AI
Message us!
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.