How China Turned 90-Second Soap Operas Into a Billion-Dollar Game Genre
You Weren't Watching Microdrama. The Audience Deciding Your Steam Launch Was.
TL;DR:
Microdrama is the largest entertainment format you have never heard of. Vertical-video soap operas, ninety seconds an episode, went from $500M in 2021 to $9.4B in 2025, overtaking China’s entire domestic box office.
It trained 830 million viewers to expect a cliffhanger every ninety seconds. Zero tolerance for filler, emotional spike every decision, screenshot-and-share dialogue. That training shows up in how they buy games.
FMV is back, and the West is not building it. Chinese FMV releases on Steam went from 3 in 2023 to 58 in 2025. Narrative is now the genre with the most hit games on the platform.
Road to Empress is the proof. $8.49, a million copies in two weeks, 96% Overwhelmingly Positive in English. The West gave up on this genre. China rebuilt it with TV-drama budgets and microdrama pacing.
The action is design, not localization. Cut filler. Treat every decision like a cliffhanger. Ship at least one shareable artifact. The trained audience does not need your game to be Chinese. It needs your game to respect the format it has been training on.
A Dead Empress Killed Me With Soup
I was three hours into Road to Empress when I realized I had been playing it wrong.
I had been treating death as failure. Reloading. Trying to find the path that kept me alive.
After my fifth or sixth death, the game’s pacing finally beat me into understanding that dying was not the failure state. It was the content.
A scholar offers Wu Zetian a bowl of soup at court. The player picks the dialogue option that accepts it.
The empress dies.
Cut to credits, a chapter unlocked, a personality reading delivered.
The point of the soup was not that you should have refused it. The point was that the scholar killed her, the scholar’s faction is now visible to you, and the next path through the chapter is a different game.
I had not been playing a branching narrative. I had been playing a microdrama with a controller.
The Western framing of Road to Empress, when there has been a Western framing at all, has been some variant of “FMV is back”. The Polygon headline was “I’m obsessed with this C-drama game that keeps killing me with soup.”
That framing is too small. FMV is a format. What is actually happening is bigger than a format.
Part 1 of this series (The China Multiplier) established the size of the opportunity. This piece explains where the audience came from.
The Format Nobody Was Watching
Western media keeps missing the microdrama because Western media keeps looking for it on Western platforms.
A microdrama is a serialized vertical-video soap opera. Sixty to ninety seconds an episode. Cliffhanger every episode. Unlock the next one with a scroll or a payment. Consumed entirely on mobile.
The format went from $500 million in 2021 to $7 billion in 2024 to a projected $9.4 billion in 2025, overtaking China’s entire domestic box office.
Read that sentence again. A vertical-video format consumed on a phone is now a larger entertainment industry in China than cinema.
830 million viewers inside China. Nearly 60% of them paying or transacting inside the ecosystem. A paying microdrama user can run up to $80 a month inside that ecosystem.
The comparison that matters is not how much they outspend a casual Steam buyer; it is that they have been trained to pay, repeatedly, for ninety-second hits of story. The wallet was already open before any game studio showed up.
Outside China, the format is projected to reach $9.5 billion globally by 2030 at a 28.4% compound annual growth rate. The United States is already the second-largest market at $819 million in 2024.
What the microdrama does to its audience is the part that matters for game developers.
It trains pacing density. Twenty episodes a day, sixty to ninety seconds each. Every episode delivers a setup, an escalation, and an emotional spike. There is no establishing shot. There is no breathing scene. The audience’s tolerance for filler has been compressed to roughly zero.
It trains cliffhanger expectation. Every episode ends on a stake. The viewer is not invited to consider whether they want to continue. The next episode is the reward for continuing. This is the same compulsion architecture as a slot machine near-miss, applied to story.
It trains the screenshot reflex. Microdrama dialogue is built to be quotable. Lines from individual episodes go viral on Weibo. The format does not just want viewers; it wants viewers who turn into distribution.
Now consider a Western narrative game with a thirty-second cutscene that exists for tonal reasons. Or a dialogue tree where two of four branches lead to the same outcome. Or a chapter that opens with five minutes of slow burn before the first decision.
The microdrama-trained player will not abandon your game over any of those individually. They will register each one as a failure of the format. By the time they stop playing, they will not be able to tell you why. They just won’t come back.
The Genre The West Killed, And What China Did With It
FMV is full-motion video. A game built from filmed footage of real actors rather than animation.
Western developers tried FMV in the 1990s. The hardware could not handle it. The acting was bad. The production budgets were small. The format earned a reputation as the worst thing about that era of gaming, and Western studios filed it under “things we tried.”
What actually happened was that the format outran the infrastructure. The reason FMV works now is that the production infrastructure caught up with the ambition, and the place where the catch-up happened first was China.
China’s entertainment industry produces television drama at a scale and quality the West has never matched.
Hengdian World Studios is the largest film and television production facility on Earth. Period costume, multi-camera 4K production, full professional casts, all available at a fraction of Western costs.
This means a Chinese FMV game is not a game studio attempting television. It is a television production with branching mechanics layered on top. The production value floor is higher than what most Western indie game studios can reach on any budget, let alone an $8.49 budget.
The numbers say what happened next:
Chinese FMV releases on Steam went from 3 in 2023 to 58 in 2025. 101 total titles across the 2023-2025 window. Narrative games now produce more breakout hits per release on Steam than any other genre, driven almost entirely by Chinese FMV output.
Love is All Around launched in October 2023 in Chinese only, with no English subtitles, and sold an estimated 1.6 million copies almost entirely within China. It topped Steam’s best-seller chart.
Road to Empress did not create the FMV audience. It was the first title to show the West what that audience had already built.
This is not a trend. This is a structural shift in the supply side of a genre Western developers wrote off.
Road to Empress, And What The Trained Audience Pays For
New One Studio in Shanghai released Road to Empress on September 9, 2025.
The team had earned credibility with The Invisible Guardian, their 2019 FMV title that picked up BAFTA recognition and 50,000-plus Steam reviews. Producer Demi Guan returned. Development consultants included Chinese history professors. The psychological profile system was co-developed with Beijing Normal University.
The setup: a fully live-action drama set in the Tang Dynasty, inspired by Wu Zetian, China’s only female emperor. Sixteen chapters. Over 100 branching storylines. Eight-plus hours of content. Over 100 ways to die. Shot in 4K at Hengdian World Studios with a full professional cast.
For Reference: Wu Zetian ruled China from 690 to 705 AD, during the Tang Dynasty, widely regarded as imperial China’s cultural and political peak. She was the first and only woman to take the title of emperor in her own right. Her story is to Chinese historical drama what Henry VIII is to British.)
The results:
A million copies sold in 13 days. 17,170 concurrent Steam players at peak.
86% Very Positive across more than 8,000 Steam reviews.
96% Overwhelmingly Positive across the 1,400-plus English-language reviews.
Number one new game on TapTap (China’s most influential third-party game distribution platform).
Number one on the WeChat popular mobile games chart at launch. Topped sales charts in Hong Kong, Macau, and Singapore.
The Price to Play ‘Road to Empress’: $8.49.
Nine months later, on June 9, 2026, New One Studio shipped Road to Empress II.
Around 1,000 minutes of 4K live-action content, roughly four times the first game’s volume. A million copies across all platforms in five days, faster than Part 1 and described by industry trackers as the fastest-selling FMV game ever released.
SEGA signed on to co-publish in Japan and select Asian regions. A Nintendo Switch version bundling both games is in development. The trained audience did not need to be convinced twice.
The Western reading of Road to Empress’ success has focused on the FMV-is-back angle. The more useful reading focuses on what design choices the trained audience rewarded.
Two of them are not obvious until you sit with the game
The psychological profile as shareable artifact.
Every choice the player makes feeds an MBTI-style personality reading. (MBTI is the Myers-Briggs Type Indicator, a personality classification framework that has become a Gen Z social currency on Chinese internet platforms.)
The reading is delivered as a shareable screenshot. Players send their readings to friends. The friends play the game to compare. The audience is now the distribution.
Cultural specificity, sharpened not softened
The Wu Zetian story is not a generic empress story dressed in Tang costume.
The game’s branching points are loaded with the actual factional dynamics of seventh-century Chinese court politics.
The historical scholars on the consultancy roster did not soften the cultural specificity for export. They sharpened it. That is what the audience came for.
C-drama is Chinese television drama, the prestige serialized kind. Road to Empress is a C-drama with a controller.
The format is exactly what the trained audience has been asking for, delivered by a studio that knew the audience was waiting.
Emotional Revenge Simulator or a Sexist Revenge Fantasy
Revenge on Gold Diggers sold an estimated 1.3 million copies in a month, and at launch it did not ship a word of English.
A Chinese-only FMV title released June 19, 2025, priced at 33 yuan, around $4.60. It was deliberately set below the market average so that, in the developers’ framing, every additional player might be one less potential victim of emotional revenge simulator which was criticized by some as a sexist revenge fantasy.
Nearly eight hours of filmed drama. 472 minutes of footage, 200-plus choices, 38 endings, one purchase, the full experience.
It hit number four on Steam’s global best-seller chart on day one, then climbed to the top of Steam’s download rankings in China, briefly outselling even Black Myth: Wukong.
Peak concurrent players: 89,127. Industry trackers estimated 1.3 million copies in June alone, roughly 73% of sales in China and 10% in the United States, all of it before any English translation existed.
On the surface it is a trashy premise. You play a man who was scammed by gold diggers and infiltrates a network to take revenge.
Chinese media and Weibo spent weeks arguing about whether it was an anti-fraud tool or a sexist fantasy.
The backlash got the creator banned from Bilibili and forced a Chinese title change to “Emotional Fraud Simulator.” It did not dent the numbers, and at launch the game ran 96% positive across more than 20,000 Steam reviews.
Underneath the discourse, the pattern is the same one Road to Empress and Love is All Around run on. Microdrama pacing. Television-grade production. A premise hyper-specific to Chinese online life, sharpened rather than sanded down. One clean purchase, no gacha, no daily login, no banner.
This is what the trained audience keeps paying for.
Love is All Around is the origin.
Road to Empress is the export proof.
Revenge on Gold Diggers is what happens when a studio takes that toolkit and points it directly at one of the most sensitive fault lines in Chinese society. They ships it with no English, survives a platform ban, and still lands top five on the planet.
What This Means Monday Morning
Microdrama trained the audience. Road to Empress proves the audience pays for the format. Everything below is downstream of that one fact: the tactics only work because the training already happened.
Here is how Western studios build for it without shipping FMV.
Audit your dialogue trees. Every line that is not advancing a stake, a relationship, or a piece of world information is costing you with this audience. The “we worked hard on this cutscene” defense is exactly the trap. Cinema-style breathing scenes do not work for a player trained on ninety-second episodes.
Restructure your consequence design. A choice that resolves immediately is a choice that did not exist. A choice that resolves three chapters later, with an emotional spike when the resolution lands, is the pattern the trained audience has been waiting for. Move your payoffs further from your decision points. The wait is the format.
Ship at least one shareable artifact. A character build worth screenshotting. A personality reading worth sharing. A custom outcome worth comparing. The trained audience is also the distribution. Games that give them something to send to friends grow faster than games that do not.
Sharpen cultural specificity, even if the culture is yours. The trained audience does not need Tang Dynasty settings. It needs settings that are something, made by people who care about that something. A Western studio making an authentically Western game with this design philosophy will outperform a Western studio making a generic action-RPG twice the size.
The Stellar Blade lesson from Part 1 was that respect for the audience moves the multiplier. The Road to Empress lesson is that the audience has a specific format in mind for what respect looks like. Most Western narrative-driven studios are closer to it than they think.
The Reframe
You don’t need to ship FMV to learn from FMV.
The format is the proof. The principles are portable. Pacing density, decision stakes, shareable artifacts, sharpened cultural specificity. Build for an audience that has been training on the most ruthless attention format in entertainment, and you will outperform a studio building for the audience the format superseded.
The audience is not Chinese. The training was.
Abbas Saleem is a Principal Consultant at Llama & Griffin, advising game studios, streaming platforms, and investment funds across six continents. He writes The Pattern Recognition: gaming industry intelligence 12 to 24 months before it becomes consensus.
Pattern recognition for game studios who would rather be early than right. How a vertical-video format nobody at GDC has watched trained the audience deciding who wins on Steam, what Road to Empress proves about design over format, and the four design moves your studio can ship without ever filming a frame.
This is the wildest Genre I never expected to get big. While I’m still trying to wrap my head around its popularity, anybody dismissing it isn’t being realistic.