Hololive: $350M in revenue for virtual YouTubers
Disney in the age of gooners
Seems like folks really enjoyed the deep dive on the Acquired podcast last week, and I’m excited to lean into more breakdowns of frontier media businesses. When I say AUDIENCE you say CAPTURE.
I don’t know how to swim, but I spent the last few days with a “descendent of the Lost City of Atlantis.” She was real in all but flesh: she sang to me, I gamed with her and her friends, and I got to celebrate her stepping into the “real” world when she performed at an LA Dodgers game. And then, abruptly and heartwrenchingly, she died—uh, “graduated”1—from streaming forever.
Pour one out for my dawg Gawr Gura.
Vtubers—”Virtual YouTubers”—are a type of content creator that overlay a 2D or 3D avatar, often with a distinct anime aesthetic, on top of their real body.2 If you squint at it, the entire industry feels like Japan’s answer to South Korea’s K-pop idol culture, where companies like HYBE have built a pipeline of young talent that dream of one day donning a manufactured persona in front of millions of doting fans. Hololive3, one of two major Vtuber labels in Japan, has effectively recreated that dynamic from first principles, albeit with an anime twist. Gawr Gura was a pop idol—her backstory was meticulously designed to balance niche and mainstream appeal, her performer was brilliant on camera, and her shark fins touched everything from music to collectible plushies.
These are overwhelmingly deep waters to wade into, and there’s only so much that makes sense to cover in this post. But it’s helpful to frame some small slice of Vtubing culture via Hololive, a mammoth in the industry that both plays a role in creating the culture and extracting from it. The nine-year-old company has evolved from VR startup to talent management agency to IP platform. They’re guiding for $350M USD in revenue in fiscal 2026. In fiscal 2025 (ended in March), their business lines broke down into:
Streaming/Content: $62M (represents 21% of total revenue and +21% Y/o/Y). This is a direct cut of superchats or YouTube subscriptions, directly from performers’ streams
Merchandising: $137M (represents 47% of total revenue, which is +64.6% Y/o/Y). This is selling products around performer IP.
Licensing/Collaborations: $38M (13.2% of total revenue and +29% Y/o/Y) Using performer IP in collaboration with other media properties, like a surprise appearance in the game “Death Stranding 2”
Concerts/Events: $52M (18% and +39% Y/o/Y). Yes, they can perform IRL.
You can see the future of the company in their revenue mix as they shift away from streaming and towards merchandising, licensing, and events. It’s a particularly interesting time for them as they transition from talent management to something more akin to Disney: an IP powerhouse.
I. Real-time anime
Anime is not a monolith, so it’s difficult to describe through a single lens. But I suspect it’s especially loved for its ability to tell stories that could not be reasonably told with live action actors. The playful animation styles can be a disarming form of escapism, but they’re just as often a Trojan Horse for real-world commentary on genocide or being an orphan. Although Vtubing aesthetically finds its roots in anime, I’d argue the artform itself might actually be an evolution of a different quirk of Japanese culture–the maid cafes that provide a socializing outlet for overworked Japanese salarymen.4 It’s widely known that while 70% of Vtube streamers are women, 82% of the audience is men.
“VTubers are, in a way, a real-time anime experience. In anime, the story is created through dialogue between the characters.” [whereas Hololive will] “set the stage and let things happen.” - Hololive CEO Motoaki Tanago
I often revisit Doug Shapiro’s post on new media businesses, where he posits that new forms of media will compete on emerging definitions of quality. You can’t gauge a new form of media by old metrics of success. It’s a helpful reframe on why Vtubing has taken off in the West over the last five years—I haven’t spent all that much time in Vtuber culture, but I don’t think I’m misrepresenting it by saying that at least some subset is looking for something more than just streaming content.5 They’re attending live concerts, they’re spending hours in live chats with their peers, and they’re dishing out tens of millions of dollars a year for “Superchats” so that their favourite streamers will see their direct messages on-stream. Notice me, Senpai.
These are men looking to belong.6
II. The residual value of digital IP
Hololive manages 89 creators, and by some estimates was responsible for 18.5% of all Vtuber watchtime last year. But talent management, especially at scale, is a brutal business. It’s both terribly low margin7 and overly reliant on big hits: you’re always worried about key-man risk when a few top performers are sustaining the rest of your talent.8
It’s a race for residuals. Subscription revenue is salary, but IP ownership is equity. The real money should come in the back-half of a character’s lifespan—Motoaki Tanigo, founder and CEO of Hololive, explicitly wants to build a “technology-entertainment company” like Disney.
One interpretation of the original Hololive hypothesis might be: what if you could build a next-gen talent management company that owns (and monetizes) its IP forever? The best case scenario would be to ride a new wave of internet-native parasociality (anime-characters-as-streamers), and then monetize on the back-end with a tried and true playbook (merchandise and licensing deals). On paper it seems brilliant: if a performer wants to retire, you would just train someone else to don their character so you can milk their merch lines til’ infinity. I’m sure Tanigo’s mouth was watering at Disney’s fiscal 2025, where Lilo & Stitch merchandise alone generated $4B of Disney’s $94B total revenue.
Hololive’s revenue mix is certainly trending in the right direction, with 41% of revenue in fiscal year 2025 coming from merchandise sales, and another 13% coming from IP licensing. But it’s not that simple.
Disney’s IP is truly disembodied from its performers. You can swap in a new voice actor, and Stitch will still be Stitch. But in Vtubing, the degree to which fans have developed parasocial relationships with their favourite performer is actually working against Hololive in this respect. They seem to follow the talent, not the character.
Take Gawr Gura as an example, my shark-queen from earlier in the post. She “graduated” from Hololive early last year, citing differences with management. A few months later, an indie streamer named Sameko Saba debuted to 1M subscribers in three days, totally unheard of for someone with no major label backing. Though it has never been publicly addressed, it’s widely accepted that this is just Gura’s original performer under a new persona. Aside from the instantly recognizable voice, the name is a tell—“Sameko Saba” translates to “little shark mackerel.” Meanwhile, Hololive seems to have ceased all sales of Gura’s merchandise. They own the IP, but it has effectively become meaningless for them.9
Merchandise is clearly a hell of a business, but it turns out that key man risk still exists when it’s rooted in human talent. Hololive needs to figure out how to maintain the value of its IP, even without its performers.
III. Come for the talent, stay for the network
Gawr Gura was one of the most successful streamers of all time, so her retirement was a material loss for Hololive. In theory, its talent development machine should be sophisticated enough to replace any major performer churn with a new “debut.” The beauty of digital IP is that more variables are in your control: you can find a hole in the market, design a particular character to fill it, train a performer, and then put the whole weight of your marketing machine to make sure that they stick the landing. But that doesn’t seem to be the case. Hololive’s last “hit” was 2023’s FUWAMOCO, and management hasn’t even attempted a new debut class since June 2024.
I suspect they’ve long known that network effects are existential for their business model. Launching new performers is a stopgap fix. It would be short-sighted to start a knife fight and just constantly race for the next breakout star against rivals like NIJISANJI. Instead, they’re structurally changing their business model. Every incremental piece of IP needs to compound in value, specifically because of its relationship to the other IP in the network. The combined effect of all the characters working in tandem has to be greater than the sum of their disparate audiences.
“The idea is to initially attract users with a single-player tool and then, over time, get them to participate in a network. The tool helps get to initial critical mass. The network creates the long term value for users, and defensibility for the company.” - Chris Dixon in “Come for the tool, stay for the network”
We’ve established that Vtuber IP is performer-dependent in a way that cartoon IP isn’t. But Disney’s IP isn’t just valuable because of its near-infinite residual value. It’s valuable because most of its IP exists alongside one another in a network, and compounds the value of the entire shared universe. Imagine trying to launch a competing superhero movie during the height of the Marvel Cinematic Universe.
Hololive has two strategies in flight to crack real network effects:
Trading Card Game
As an outsider, it seems like a successful Trading Card Game (TCG) is the holy grail of a subculture. Think Pokemon or Yu-Gi-Oh. It’s at once a collectible, a community event, and a secondary market of merchandise that appreciates over time. So it makes sense that Hololive made a big push to launch their own TCG in September 2024. It’s a structural advantage for every part of the business:
Increase the IP longevity of new talent. In theory a Gawr Guru trading card would continue to be valuable in the context of the game and increase in valuable as a collectible
Disincentivize existing talent from going indie. There’s a lot of existing infrastructure to spin up a merchandise line for yourself, but a TCG by definition requires a vast network
Spin up a lucrative new merchandise line to supplement the rest of the business. As a benchmark: some reports peg Yu-Gi-Oh’s physical TCG at $302M revenue in Japan alone for fiscal 2023.
Early signs are looking good for the TCG, clocking $13M USD revenue from April-June 2025 (33% of fiscal Q1’s total merchandising revenue).
Holoearth
Now that the first Vtuber skin has arrived in Fortnite, the starting pistol has been fired to see if gaming will be a meaningful new revenue line for the industry. Holoearth, Hololive’s gaming metaverse project, launched in beta in April 2025. I played it for an hour, and it was pretty underwhelming. Where Fortnite had the advantage of being a hyper popular shooting game before it leaned into its metaverse-like events, Holoearth is racing to do too much at once. It’s at once a survival game, similar to Minecraft, and has metaverse components like concerts and avatar marketplaces. But remember that Holoearth was originally a VR company. This feels more like a return to form than an attempt to build a brand new core competency, so I’m sure they’re just getting started here.
Hololive is a sprawling business, which happens to exist in the sprawling subculture of Vtubing, which also then happens to exist in the sprawling subculture of anime fandoms. There’s only so much I can cover in this one post, and I suspect I’ll be revisiting this world often. But it’s a fascinating study on how technology can create truly novel cultural artifacts—Vtubing would not exist in its current state if not for a fortuitous intersection of streaming, pandemic-era lockdowns, and generational strides in avatar mapping technology. But culture is not a business, as any blogosphere-era writer will tell you. And that’s fine and special on its own! But a culture that is supported by an underlying business model is something special indeed. You can look around at Substack and catch a whiff of the same.
If we take my earlier arguments at face value, that this is largely a subculture of young men who are just looking to belong to something, then it is equal parts heartwarming and gut-wrenching that this shared universe exists at all. I’m excited to see where this world evolves from here, and I do hope that Hololive finds a way to win.
Industry parlance for a performer permanently retiring their character.
It’s amazing that some people in the Vtuber community will refer to human streamers as "fleshtubers".
Hololive’s parent company is publicly traded on the Tokyo Stock Exchange as Cover Corp.
Remember that anime is no longer niche. According to Dentsu, a Japanese ad agency, 36% of Americans watch anime at least weekly.
According to this YouTube study, “55% of viewers agree that their content choices collectively create a sense of belonging.” Tbh the study seems laughably weak, with something like 600 survey participants specifically in Italy. But, intuitively, I think the point stands.
When a fan sends money directly to a streamer on YouTube or Twitch, Hololive counts that as subscription revenue. The exact splits are unclear, but platforms take 30%, and then, by most estimates, Hololive pays the streamer 50% of that net amount.
Hololive experiences an average of 1-2 “graduations” a year, where a performer decides to step away from streaming for a variety of reasons. But there’s a worrying trend here, with over 6 graduations in 2025 alone.
In September 2024, Hololive launched an “affiliate” program to combat this. The idea was to let performers step away (permanently or temporarily), but still work together on merchandising or sporadic events. It doesn’t seem to be working so far, with only two of 2025’s graduations opting for the affiliate model instead of a clean break.
"When I say AUDIENCE you say CAPTURE." lol