Two ways to live in short drama: scale protects the capital, content runs on relationships
Short drama has two survival modes: the scale mode runs on capital safety and the near-certain returns of 300 titles a month, the content mode on character relationships that determine motive; both sides think AI cannot replace a live actor's micro-expressions for now.
The video won't play here. Listen to the audio instead:
The argument · tap a timestamp to hear it
Xi'an and Zhengzhou won short drama because Hengdian's studios face the wrong way
The backlots at Hengdian (横店) were built for horizontal, deep-focus shooting, so converting them costs more than working in Xi'an (西安) or Zhengzhou (郑州). Short drama was commercialised into existence in Xi'an because, at the tail end of the pandemic, Xi'an was a tourism city whose real economy needed a restart, and its culture-and-entertainment workers embraced short drama more willingly. What Zhengzhou and Xi'an have in common: easy transport, a population above 13 million, an existing performance-marketing gene, a large pool of cheap labour, and people willing to try something new. Zhengzhou is more cut-throat than Xi'an, with a more complex industrial structure and a richer roster of leading company names.
— Hou ChaoOne city's ceiling is about 300 titles a month, and coordination is why
At the peak, 200 projects were shooting simultaneously. Counted against the number of people working the key skilled roles in Zhengzhou alone, full capacity comes to roughly 300 titles a month. Short drama is a coordination-heavy craft: the more projects run at once, the more the scheduling load multiplies, so 300 is the city's carrying limit — and before this, no company in the industry had been able to manage one or two hundred projects at the same time. What holds that pace up is a script-IP library, genre tagging, heavy reuse of the same directors and similar locations and props, and central-office control.
— Hou ChaoGoing free replaced the hard hook with completion rate, and she resisted it
The biggest change in moving from paid to free was the metric: completion rate took over from the hard hook. Li Jiajia (李佳佳) resisted at first and felt inadequate about it, believing her own method could not keep up with the rhythm of free. Only when Shijian Jie (时间节) became a hit on the paid side and then on the free side did it confirm that the character-relationship logic holds on the free end too — it became a milestone for the team. From then on the method shifted from concentrating effort on the first ten episodes to spreading it relatively evenly across every episode and simply telling the story well.
— Li JiajiaCharacter relationships are the foundation; flat archetypes are why stories fail
The methodology is ‘character relationships determine motives, motives determine events, events determine the story’. What makes the hit Su Taitai Gaodiao Lihun Le (苏太太高调离婚了) work is that each character stands at a different distance from every other, so the tone they speak in and the information they let slip differ accordingly — solid detail, the opposite of flat archetypes. A story is unwatchable precisely because its characters are archetypes. Ensemble scenes are the hardest to shoot, because every line a person says has to carry a different attitude toward every other person in the room.
— Li JiajiaAdmit the point of the company first: it is to make money
One guest says frankly that he ran a deep breakdown on himself and concluded that his real goal is to make money. The company's second boss, a super-achiever, handed him ‘a small target’ so he would keep earning; he found he could not accept throwing away his experience, his network and his team, so he keeps going to work, making money, making shows — a loop. He also judges that in long-form film and TV not one company has sustained itself on the strength of good content, unless the boss himself is the content person.
Batch shops review for fatal flaws; premium shops review for the ceiling
On Hou Chao's (侯超) side, out of a 300-person content team only two or three people can actually put 3-5 million yuan behind a single project, so his post-mortems look only at ‘fatal factors’: genre timing, whether an ordinary person can execute it, the difficulty of shooting logistics — the lethal shortcomings. Once identified, a notice goes out banning a repeat. Li Jiajia's projects more often fall short of expectations without dying outright, so she analyses the ceiling from the production end and the topic-selection end; the two ends can compensate for each other, though production usually cannot patch a hard flaw in topic selection. A runaway hit always comes from a long pole: only a show whose strength is spectacular enough can blow up.
AI and live-action are becoming two separate lanes, not one contest
AI and live actors are already two lines: AI is better at rendering spectacle people have never seen but can imagine — Shanhaijing (山海经) creatures, effects, animals; live-action fits emotion and household-intrigue material that rests on relationships. A live actor's micro-expressions and the tension a good actor carries are things AI will not learn any time soon, and as long as that difference exists, live-action has room to survive. The defensive logic on the other side is to require only that a title clear the ‘kill line’ — no single title loses money — protecting the floor while chasing the ceiling. AI has also already landed in script tools, electronic script-supervisor sheets and overload-proof scheduling, not just as content assistance.
Interactive film games are not too hard, they are too unproven
Facing the rush into interactive film games, one guest says he seriously considered dissolving the company and starting over once AI arrived, because of path dependence. But the reason he will not make an interactive film game like Shengshi Tianxia (盛世天下) is not difficulty: it is that no demo has genuinely worked yet and the industry's mindshare and rules are underdeveloped, so he is unwilling to be the first to try the crab. The other adds that the newer the company, the thinner its talent bench, and nothing is more damaging than split attention — ‘want to do everything, get nothing done’ — because the opportunity cost of the boss's and the partners' time is too high.
In their own words · checked verbatim
First is capital safety. Because you only get to earn anything if you don't lose money first. Second would be, uh, our profit at scale.
第一是本金安全。 因为先不赔钱才有的赚嘛。 第二就是属于是我们的呃利润规模。
Hou Chao8:07
If I had to sum it up, everything the audience can't see and can't feel has to be cut.
如果说总结的话，其实一切观众看不到的，感受不到的都要省掉。
Hou Chao18:05
Who doesn't want to make that kind of good thing? Who doesn't want to polish something carefully, something that everyone both praises and pays for.
谁不想做这种好的东西呢？ 谁不想精心打磨，做一个大家都又叫好又叫做的东西呢。
Li Jiajia35:48
The best director, the best screenwriter, the best actors, the best script — but added up, not the best work.
最好的导演，最好的编剧，最好的演员，最好的剧本，但加起来不是最好的作品
Hou Chao39:59
The core of acting is that the actor who fits this role best goes and plays this role well.
演戏的内核就是最适合这个角色的演员过去演好这个角色
Li Jiajia42:05
I think I can write something like Xuezhong Handaoxing (雪中汉刀性). I think that achievement in this lifetime, wow, should make me happier than earning 10 billion.
我觉得我能写出来雪中汉刀性这样的内容。 我觉得这辈子的成就哇比比应该是比我赚100个亿都要开心。
Hou Chao1:09:08
Figures
| Monthly capacity ceiling for live-action short drama | about 300 titles per month (estimated from one city's carrying capacity of working professionals) | 5:17 |
| Cost per title, batch approach | 200,000-250,000 yuan | 15:36 |
| Cost per title, premium approach | 400,000-800,000 yuan | 16:04 |
| Screenwriters on staff at the peak | three to four hundred | 7:22 |
| Live-action short dramas platforms have planned for the second half of the year | over 10,000 | 1:07 |
| Sales of the interactive film game Shengshi Tianxia (盛世天下) | over 5 million copies | 1:02:27 |
Glossary
- RY (input-to-output ratio, return multiple)
- The short-drama industry's metric for a project's profitability: an RY above 1 means profit, and Hou Chao uses it to decide when to greenlight titles in batches.
- DIT (Digital Imaging Technician)
- The role that copies and transcodes footage on set and manages the script-supervisor sheets; expensive and slow, and Hou Chao has replaced part of it with AI.
- Zhanshaxian (斩杀线), the kill line
- The break-even point for a single title; clearing the kill line means you get to keep going, and it is the minimum survival standard of the batch approach.
- Jinshouzhi (金手指), the protagonist's cheat power
- The outsized ability or system that only the lead has in web fiction and short drama — X-ray sight, a treasure bowl, gambling on rough stones — a stock ingredient of hit genres.
How to listen
Short-drama writers, producers and investors most of all; also useful for anyone following AI content startups or trying to understand how scale and craft can be had at the same time.
If you already know the industry, skip the opening rundown of platform policy (0:00-2:10); the costume-drama chat at 12:43-13:00 can be fast-forwarded.