Why Is the *Kung Fu Girls' Football* Marketing Push So Perfectly in Sync? I Used AI to Trace the Money Behind All 24 Production Companies

The marketing for Kung Fu Girls’ Football (功夫女足) over these two weeks has been eerily synchronized, wave after wave. I haven’t seen the film and I’m not going to say whether it’s good—what I couldn’t stand was how in-sync it all felt, and I wanted to figure out where that rhythm was coming from.

Kimi K3 had just launched, and I’d been wanting to test how it performs in real work behind WorkBuddy, my AI digital employee. Two birds, one stone: work through the talking points on one track, and drill through all 24 production companies behind this film on the other. WorkBuddy does the work, K3 is the model under it, and Qichacha’s official data feed supplies the facts—shareholders, ultimate controllers, registration changes, share pledges, all from the registry, not from what some influencer said. In one afternoon it produced a full research report as a web page. This job used to take a financial journalist a week.

Two counterintuitive conclusions up front

One, no matter how loud the criticism gets, the box office will be fine, and that has very little to do with whether the film is actually good. Two, by the end of the drill-down, a name surfaced that has no business being on that list at all.

The talking points: a ladder that climbs from your age to your country

Before release, the pitch was nostalgia—“we still owe Stephen Chow a movie ticket,” a decade-old line brought all the way back. Buying a ticket got repackaged as settling a debt, as saying goodbye. On opening day Douban came in at 6.6, and every talking point after that was a counter to a specific complaint. Say it’s not good? Then you’ve just grown up. Say it’s not funny? The kids were loving it. Say the plot is weak? Then you missed how hard everyone worked. Notice something? Not one of those four lines actually addresses whether the film is good. Every one of them switches lanes.

Then the ladder gets one more rung, the highest one—nation. Analysis pieces comb frame by frame for metaphors: the rival Dahe team is a homophone for Yamato, and the eight-year run of a seven-time champion getting toppled maps onto the eight-year War of Resistance. In the last couple of days it went further. Large numbers of seemingly ordinary accounts started running the same script: a supporting character speaks for Iris Chang, the opening strafing scene honors the fallen. Climb to that rung and a bad review becomes a question of allegiance, essentially unarguable. Honestly it leaves me at a loss. That history carries far too much weight to be used as a shield against bad reviews. No movie has earned the right to it.

The real offensive is making enemies and building a faction—box office theft accusations, an imagined opponent getting swarmed—while the person saying the least of anyone is Stephen Chow himself. The shouting matches aren’t a marketing accident. The shouting matches are the marketing engine. Every round of argument moves the box office up another step.

Following the money: the ticket sellers, the score setters, and the trend makers are all on the same side

You can’t see the conductor by looking at content. You have to follow the money. After drilling through all 24 companies, the answer is more sophisticated than any guess online.

• Maoyan, the app you buy your tickets on, is both a production company and the lead distributor for this film. The 9.4 opening score is one it set itself, while Douban has it at 6.6—the referee suited up and joined the game. Drill one layer deeper and the largest shareholder bloc in Maoyan’s domestic entity is the Enlight camp, roughly 47% combined.
• The operating entities behind Weibo and Bilibili are both on the production company list. Douyin has no equity stake, but Bingo Group signed an AI partnership with Volcano Engine, so the ecosystem connection is there.

Do you see the net? The ticket sellers, the score setters, the trending-topic makers, the fan-edit creators—all on the same side. Do you still need a director-in-chief issuing orders one by one? You don’t. There’s a rhythm, but no command center anywhere in the records—the structure of the money is the baton.

The hidden winner off the list

The name Ruyi does not appear on the production company list, yet multiple financial outlets are writing about China Ruyi’s Kung Fu Girls’ Football. Half of the 380-million-yuan budget went into VFX, all of it handled by its own in-house AI visual system. Nearly 200 million yuan in VFX fees was banked before release, so how the box office performs doesn’t affect that money at all. And Wanda Cinemas, which it controls, takes a cut of every yuan the box office collects. No such name in the records, yet it controls where half the budget flows. It didn’t bet on the box office. It bet on the supply chain.

Chairman Ke Liming spent five years at a Hong Kong hedge fund before moving into film and television—finance first, content second, and that order determines his entire playbook. Three separate moves, all the same maneuver: sell assets high, pick up control low. I also had AI run a risk scan on him across 18 categories. Seventeen came back at zero, unnaturally clean for someone in this industry.

The conclusion: what’s scarce isn’t conclusions, it’s verification

Around any hot topic, false information doesn’t show up in scattered bits. It’s mass-produced. In that environment, conclusions aren’t scarce. What’s scarce is the verification standing behind a conclusion. This kind of drill-down capability used to belong to financial journalists and due diligence lawyers. Now, one digital employee, one domestic model behind it, one public data feed, and it belongs to anyone who knows how to ask. Apply it to your own business—a competitor, a supplier, a partner you’re about to sign with. Who’s behind them, where the money comes from, whether there are traps. One afternoon covers all of it.

Next time you see a perfectly synchronized rhythm, don’t rush to pick a side, and don’t rush to attack. Go check who’s standing behind the rhythm first.

This is the full transcript of this episode. Timecodes were generated by AI aligning to the audio, for search and citation. This episode features an AI digital avatar (likeness and voice trained on my own material).

[00:00] I had Kimi K3 work through the whole capital structure and the whole media narrative around Kung Fu Girls’ Football. Let me be clear up front: I haven’t seen the film, and I’m not going to say whether it’s good. What I couldn’t stand was the wall-to-wall marketing these past two weeks—wave after wave, with a rhythm so tight it was scary. Honestly, it made me sick. And I got curious: where does that rhythm actually come from? K3 had just launched, and I’d been wanting to test how it actually holds up paired with WorkBuddy on a real job. Two birds, one stone.

[00:23] Two things. I worked through all the talking points flying around. And I ran a check on all 24 production companies behind this film, one by one. Once it was done, my biggest takeaway was just one thing. With AI, for the first time, ordinary people like us can watch public opinion get steered start to finish and see it plain as day—including who’s standing behind it counting the money. For this film specifically, two conclusions. One: no matter how loud the criticism gets, the box office will be fine, and that really has very little to do with whether the film is any good. Two: by the end of the drill-down, a name surfaced that has no business being on that list at all. I’ll get to that name later.

[00:51] Let me quickly explain how I did it, so you can do it yourself after listening. The one doing the work is WorkBuddy, my AI digital employee, and the model under it this time is K3. Then I wired in Qichacha’s official data feed. Shareholders, ultimate controllers, business-registration changes, share pledges—all of it from the registry, not from what some influencer said. And while it worked, it cross-checked against listed companies’ public filings. When it finished, it didn’t dump a pile of chat logs on me. It output a research report as a web page—the one you’re looking at on screen right now. Timeline, ownership structure, a map of the talking points, all in there.

[01:16] One afternoon, 24 companies drilled through. This job used to take a financial journalist a week. The skills and prompts I used this episode are at the end of the video—but let me keep going first. Start with the talking points. I’ll lay out what came across my feed, and you tell me if it looks familiar. Before release, the play was nostalgia. “We still owe Stephen Chow a movie ticket”—that line goes back to 2016, around The Mermaid, and this time it got fully revived. Then a batch of articles titled “Goodbye Stephen Chow, goodbye youth,” saying this might be his last one. The act of buying a ticket got repackaged as repaying a debt, as a farewell ceremony.

[01:44] And it wasn’t just marketing accounts running with it. Stephen Chow himself picked it up at a promo event, saying nobody owes him a ticket, that if anything he feels embarrassed. The moment he said that, another round of trending topics. A ten-year-old joke among fans, and once the man himself picks it up, the joke goes official. The nostalgia track went from marketing accounts making a meme to something mutual. On opening day the criticism came out, and Douban opened at 6.6. Then look at the talking points that showed up next—every single one of them is a counter to a bad review.

[02:07] Say it’s not good? Then you’ve just grown up, short video rotted your brain. Say it’s not funny? The kids were having a great time, and Chow’s only job is making people happy, is that not allowed? Say the plot is weak? Chow poured his heart into it—the shoes are an easter egg from 25 years ago, the empty chair is a tribute to Ng Man-tat—and the think pieces come one after another, so if you can’t see the deeper meaning, that’s your problem. And if that fails, go straight to values: a Stephen Chow film was never just a film. Notice something? Not one of those four lines actually addresses whether the film is any good. Every one of them switches lanes. You talk quality, it talks about your age. You talk plot, it talks nostalgia. You say it’s not funny, it says you’re the wrong audience. Four channel changes, and your bad review hits nothing but air every time.

[02:42] And just in the last couple of days, the ladder got another rung—the highest one: nation. The analysis pieces started combing frame by frame for metaphors. The rival Dahe team is a homophone for Yamato. They’re seven-time champions toppled in year eight, which maps onto the eight-year War of Resistance. The tiger in the stands stands for the imperial powers. Even the final winning push gets read as the last breath spent right before victory. Follow that reading, and not getting it isn’t just you having grown up anymore—now it’s you not knowing your own history. There are even articles flat-out calling the outlets that ran bad reviews a coordinated hit job, saying that under the comedy shell there’s a patriotic backbone nobody gets to smear.

[03:13] And in just these couple of days the reading keeps escalating. Large numbers of accounts that look like ordinary people are running the same script: the supporting character Sang Biao is speaking for Iris Chang; that strafing scene at the opening is honoring the fallen. Scroll through them one by one and the wording is nearly identical. Follow that line and buying a ticket and walking into the theater becomes remembering history, honoring the martyrs. At that point, saying the film isn’t good isn’t a matter of taste anymore—it’s a matter of allegiance. And there’s almost no arguing with that one. Look at the ladder: from age, to audience, to nostalgia, all the way up to nation. Once it climbs to that rung, a bad review has nowhere left to land. You’re reviewing a film, and what it answers you with is a chapter of history.

[03:46] Honestly, this is the part where I just don’t know what to say. The history Iris Chang recorded, and those who died—that weight is far too heavy. Using all of that as a shield against bad reviews for a movie? The movie doesn’t come close to deserving it. Okay, all of that is still defense at the level of talking points. The real offensive is making enemies and picking sides. First came the box office theft. Photos of handwritten tickets everywhere, the studio putting out a statement saying it would cut the projection keys if confirmed, Chow himself posting one of those blank slips with three question marks. Sympathy maxed out—buying a legitimate ticket now equals protecting Stephen Chow.

[04:09] Then came the imagined enemy. Two TV hosts said publicly that the film was bad. One had their comment section swarmed shut the same day; the other had a six-year-old remark dug up and took tens of thousands of attacks. Notice this: the person who said the least of anyone was Stephen Chow himself. Twice in a whole week—once with that “embarrassed” line, once with those three question marks. Both times it trended. The act of hitting back got done for him, by the community. Watching this movie, at this point, stopped being about whether it’s good and became about whether you stand with Stephen Chow.

[04:34] The slickest move was swapping the print. On day four they officially announced a new version, with over a thousand VFX fixes. Sounds like craftsmanship, right? But check the dates. The day the new print went live was exactly this film’s originally scheduled release date. What does that mean? To grab the summer window, it opened six days early. What the first wave of audiences watched was a version that hadn’t reached its own delivery date. The way I read it, a half-finished product wasn’t an accident—it was a price already agreed to back when they went for the window.

[04:58] And that print swap killed three birds with one stone. Every technical complaint from before is void—you were trashing the old version. Repeat viewings got a new reason: the new version is different. And the opening-day old print? It turns into a collector’s item. A product-quality accident got operated into a three-way win. And go line it up against the timeline: every round of shouting moves the box office up a step. The day the host started trashing it, opening day crossed 100 million. The old remark got dug up and swarmed—600 million in three days. Three question marks go out—past a billion. The shouting matches aren’t a marketing accident. The shouting matches are the marketing engine.

[05:23] Okay, talking points done. So here’s the question: with a rhythm this tight, who’s conducting behind it? You can’t see it from the content alone. You have to follow the money. And that’s what WorkBuddy did once it was wired into the Qichacha feed. After going through all 24, the answer surfaced—and it’s more sophisticated than every guess online. What surfaced is a net like this. Company one, Maoyan. Yes, the Maoyan you buy your movie tickets on. It’s a production company on this film and the lead distributor. That 9.4 opening score? Set on its own platform. Douban has it at 6.6. And Maoyan Pro, the tool the whole industry uses to watch screening counts and forecasts, is also theirs. What does that add up to? The referee walked onto the field and started playing.

[05:58] Drill one layer up and it gets more interesting. Guess which bloc is the largest shareholder in Maoyan’s domestic entity? The Enlight camp, roughly 47% combined. And how did Enlight Media answer the exchange? The company did not participate in this film. At the level of direct investment, that statement is fine. But it’s the largest shareholder bloc in Maoyan, the production and distribution company. The space between what they say and what’s actually true—that’s exactly where it hides. Company two, Weibo. One of the co-production companies is called Beijing Weimeng Chuangke. The name may not ring a bell—it’s Weibo’s operating entity. Which means for every single trending topic you see, the platform itself is an interested party.

[06:28] Company three, Bilibili. One of the production companies is called Wanwei Renhe, and once you check the equity, Bilibili’s operating entity Shanghai Huandian holds 9%. Company four, Douyin. There is no ByteDance among the 24—not one—and that was confirmed company by company. But Stephen Chow’s Bingo Group signed an AI partnership with ByteDance’s Volcano Engine, and the whole main cast rallied on Douyin to build hype. No equity, but ecosystem.

[06:50] So now, do you see the net? The ticket sellers, the score setters, the trending-topic makers, the fan-edit creators, the traffic pumpers—all on the same side. Even the influencers don’t need anyone to direct them. “Actor in a Stephen Chow film”—that phrase alone is the highest-grade endorsement an MCN can get, so they’ve got ten thousand reasons of their own to jump in. So does that whole set of talking points still need somebody up top handing out orders one by one? Not anymore. The structure supplies the incentive, the algorithm amplifies the conflict, and the community brings its own ammunition. There’s a rhythm, but no command center anywhere in the records. The structure of the money is the baton.

[07:18] That’s the first layer of why the box office will be fine. Before this film ever opened, it had turned its distribution channel, its rating gateway, and its topic platforms all into its own shareholders. Its word of mouth can tank, but it cannot go unnoticed. That said, this visible net isn’t the most interesting part. Remember what I said at the top—there’s a name that isn’t on the list at all, but it might be making the safest money in this whole game. Let’s talk about it now. During the cross-checking, it caught one anomalous signal. The word Ruyi does not appear on the production company list. But multiple bylined financial outlets are all writing about China Ruyi’s Kung Fu Girls’ Football.

[07:45] Put the pieces together and here’s what it is. Of this film’s 380 million yuan budget, half went into VFX. All 1,200-plus VFX shots were handled by China Ruyi’s own in-house AI visual system. And per reports, it also had a piece of the investment indirectly, through a subsidiary. No such name in the records, yet it controls where half the budget flows. Then look at how it takes its money and you’ll see what a different league looks like. Nearly 200 million yuan in VFX fees was already banked before release, so how the box office does has nothing to do with that money. Wanda Cinemas, which it controls, is the largest exhibition network in the country, and for every yuan the box office collects, the exhibition end takes a cut for it. And its own Nanguadianying streaming service holds the streaming rights on the back end.

[08:14] The platforms get paid for traffic. Ruyi gets paid on the entire supply chain. I ran a rough number. On VFX and exhibition alone, even at the current box office, the Ruyi side is already looking at somewhere around 200 million yuan coming in. If the box office really gets to 3 billion, that number climbs to something on the order of 500 million. And note: most of that money has nothing whatsoever to do with whether this film makes or loses money. So everyone else is betting on the box office. It isn’t. The bet it placed isn’t on the box office—it’s on the supply chain. So why not put its name on it? The way I read it, the upside of a credit is vanity, and the risk is getting called the capital that hijacked Stephen Chow. Don’t take the credit, and you catch none of the abuse and lose none of the money.

[08:44] By this point my question had completely changed shape. Who built a setup this big? China Ruyi’s chairman is Ke Liming. From Huanggang, Hubei, born in the 1980s. This name’s presence in mainstream media is wildly out of proportion to the cards in his hand. Wanda Film already changed its name back in April this year—it’s Ruyi Film and Entertainment now, and he’s the ultimate controller. The largest cinema chain in China now carries the Ke name. There’s one key detail in his résumé. Before he got into film and TV, he spent five years at a hedge fund in Hong Kong. Finance first, content second. I think that order is what determines his entire playbook.

[09:11] He was lead investor on Hi, Mom and Moon Man, and a co-investor on Lost in the Stars—a hit rate that’s top of the industry. But what’s really worth talking about is three moves he made. Three times, all the same maneuver: sell assets high, pick up control low. Jump one, 2020. He sold Ruyi Pictures plus Nanguadianying, which he held, for 7.2 billion Hong Kong dollars to a listed shell jointly owned by Evergrande and Tencent. Disclosed net profit on those assets the prior year: somewhere around 40 million. And the timing of the deal? The eve of Evergrande blowing up.

[09:36] Jump two, 2021. Evergrande was in a rush to cash out and get out, so he turned around and put up over 4 billion Hong Kong dollars to take the shares Evergrande was dumping. Evergrande is out, the shell gets renamed China Ruyi, and it’s his. Look at this stretch: the money that bought the assets came from the Evergrande side, and he stepped in to take over at Evergrande’s most desperate moment. The only winner in this stretch is him. Jump three, 2023. Now it’s Wang Jianlin who’s short on cash. In two tranches, roughly 4.4 billion yuan total, he swallowed the controlling stake in Wanda Film. Per disclosure in Southern Metropolis Daily, 60% of the second tranche, just over 2.1 billion, was funded by loans.

[10:06] The structure of that third jump is the most elegant thing I saw all afternoon. The funding timeline it drilled out looks like this. July 4, 2023: Tencent injects 4 billion Hong Kong dollars into China Ruyi and becomes its largest shareholder. July 28: the domestic acquisition vehicles under Ruyi have over 4 billion yuan paid in. Same month, the Wanda deal closes. Three weeks end to end, airtight. The ownership structure is even better. The acquisition vehicles are two small companies, registered capital of 3 million for one and 10 million for the other, both 99% personally held by Ke Liming. What does that add up to? The control chain over China’s largest cinema chain hangs off two tiny shells capitalized in the millions, and sits entirely outside the listed company where Tencent is the largest shareholder. China Ruyi even put out a specific announcement: no consolidation, no board seats, no involvement in day-to-day operations.

[10:40] Let me be fair here. Half of that structure was forced by regulation. The negative list for foreign investment still, to this day, bars foreign capital from film production, distribution, and cinema chains. A Hong Kong-listed company counts as foreign, so onshore this is the only way to build it. This is a standard compliance move across the industry, not a conspiracy. But the effect is very real. Tencent’s money gets the returns from the ticket pool, while the sentence “Tencent controls China’s largest cinema chain” doesn’t hold up under any regulatory definition. For a Tencent that’s in the spotlight, that effect is exactly what you want.

[11:03] I also had it run a risk scan on Ke Liming himself across 18 categories. Credit blacklisting, enforcement actions, travel restrictions, tax violations—17 of them came back at zero. Unnaturally clean for someone in this industry. So what kind of figure is Ke Liming? The way I read it, nobody at this table does the math better than he does. Everyone else is doing math on box office; he’s doing math on structure. Three moves, and every time it’s the same thinking: keep the certainty for yourself, leave the risk to others. Look back at how Ruyi entered this film—exactly the same, isn’t it?

[11:25] There’s exactly one place in this whole case I couldn’t get to the bottom of: who actually made that 2.1 billion loan. Public information stops there. If somebody in the comments knows, point me in the right direction. Last minute—back to the part that’s actually useful to you. Everything just now, the equity, the pledges, the change records, the paid-in dates, all of it comes from public business registrations and listed company filings. Anyone can look it up. What WorkBuddy did was wire together public data scattered across dozens of places, line it all up, filter out the fake stuff, and then lay it out for me as a web report I can just read.

[11:49] Oh, and the test I mentioned at the top. This was K3’s first real job since launch, and you’ve already seen how it holds up. In that one afternoon alone it knocked down a whole string of false claims along the way. The rumor that Alibaba, Enlight, and Wanda co-produced it—all three listed companies answered for themselves: not involved, made up by marketing accounts. On the day the rumor said box office had hit 2.5 billion, it had actually only just crossed 1 billion. One company on the list was rumored to be part of the Alibaba camp; check it and it’s a company owned by two individuals, purely riding on a similar name.

[12:12] Notice something? Around a hot topic, false information doesn’t show up in scattered bits. It’s mass-produced. In an environment like that, conclusions aren’t scarce. What’s scarce is the verification standing behind a conclusion. This kind of drill-down capability used to belong to financial journalists and due diligence lawyers. Now: one digital employee, one domestic model behind it, one public data feed, and it belongs to anyone who knows how to ask. Apply it to your own business—your competitor, your supplier, the partner you’re about to sign with. Who’s behind them, where the money comes from, whether there are traps. One afternoon covers all of it.

[12:37] As promised at the top, the skills and prompts I used this episode are in the next few frames, exactly as they are—pause and take them. And next time you see a perfectly synchronized rhythm, don’t rush to pick a side and don’t rush to attack. Go check who’s standing behind the rhythm first.