09. After Apple, Fewer Brands Believe They Can “Change the World” | Dreame Drops In to Address Three Major Critiques
Summary
- Dreame’s truly reusable moat is not a single motor specification, but an integrated system spanning R&D, domestic component sourcing and in-house production lines. The company entered high-speed motors around 2015 and, in roughly 1.5 years, pushed the reference line from 20,000-30,000 rpm for domestic suppliers and 100,000 rpm internationally to 125,000 rpm; 郭人杰 says Dreame subsequently became the world’s first developer to achieve 150,000 rpm. At the time of recording, it was mass-producing 180,000-rpm motors, had 200,000 rpm in reserve and had turned its first roughly 50-person production line into a lights-out factory. 郭人杰’s core view: “We’re confident that, with R&D and the supply chain, we will make it to No. 1.”
- Seven years of technical accumulation converted rapidly into revenue after 2020, but the group and China-business figures need to be separated. The China business went from zero to RMB300M in 2021 and RMB3B in 2022, with a target of roughly RMB6B in the year of recording; the company-wide figure cited at the opening of the program was close to or had reached RMB10B. The explanation was not “astonishing marketing ability,” but the fact that while others needed 6 months to launch a product, Dreame could bring one to market in “3.5 months” on the back of its technology and supply-chain reserves.
- Dreame chose high prices not to wear a premium label, but to close the loop of high R&D investment, leading products, high profits and further R&D. 郭人杰 asks: “Why should the best product in the world be sold at a lower price?” Dreame says it entered China, Germany, France and Italy at the highest price points and won those markets; a vacuum priced above Dyson’s instead posted the lowest return rate in company history, suggesting that price-insensitive users cared more about “the best product experience in the industry.”
- Dreame answered accusations of being a “capital game,” unstable and lacking substance with a four-way allocation of its 12% R&D investment. The 4 buckets of 3% cover all existing industry technologies, further performance improvements, cross-industry innovation and future categories such as robotic lawn mowers; 郭人杰 says outsiders “mistook the last 3% for the entire 12%,” while the first 9% still supports the existing business. This is the company’s own framing, not proof that the criticism has been fully disproved by financial data, but it does provide a clear resource-allocation logic.
- Dreame did not use a price war to force a No. 1 position in any single category, choosing instead to stay mid-to-high end and accept a division of labor in which “some companies do technology while others do mass affordability.” At the time of recording, it ranked No. 2 domestically in both wet-dry floor cleaners and robot vacuums; its high-end robot-vacuum share was about 60%, versus roughly 20% across the full market, while the No. 1 player was about 1.5x its size. 郭人杰 says Dreame “could have a chance to become No. 1” by cutting prices and raising marketing spend, but that would damage its own and the industry’s business model.
- The key to Douyin growth was not unlimited spending, but turning products, audiences and demonstrations into high-frequency data experiments. In 2021, the team caught Douyin e-commerce’s shift from white-label goods to brands and used “vacuuming instant noodles” to turn a wet-dry cleaner’s complex specifications into an instantly legible household disaster; daily sales rose from RMB200,000 to RMB1M. A three-dimensional combination of front-end content, livestream demonstrations and audience targeting could identify a direction within 1 week. After the traffic dividend disappeared, the company says it pushed its Douyin ROI requirement to 1:10, versus an industry average of roughly 1:2.5: “In the end, all that’s left is organizational efficiency.”
- The real improvement in growth quality came from shifting focus from the 90% who had never used a robot vacuum to the 10% installed base, using tangible innovation to repair the category’s reputation. The X20’s mechanical arm addressed the known problem that robot vacuums could not clean roughly 5 cm along walls; it won out from 37 designs, sold for RMB5,299 when the industry average was RMB3,999, broke out immediately after launch and quickly became No. 1 in Germany. Dreame thereby achieved what it calls the impossible triangle of higher prices, lower expenses and scale growth. Its next step, previewed during the recording, was a mechanical leg able to clear a 5-6 cm threshold and solve another familiar failure point for existing users.
- “Becoming a company more formidable than Apple in 5 to 10 years” is Dreame’s most conspicuous ambition, and it comes with the execution costs of a young organization. A team averaging roughly 26-27 years old believes “the world is a makeshift operation” and “either don’t do it or be No. 1,” compressing the time needed to admit mistakes to a few seconds. The costs include a data-first elimination mechanism that damages the employer brand and a supply-chain accident after Douyin took off in 2021: inventory sold out in just 3 weeks, and replenishment took months. Whether the ambition holds will depend on whether that speed can be converted into products and brand equity over the long term, while the organization’s holes are filled in time.
Deep dive
1. An Aerospace Club’s Technical Base Ultimately Became a High-Speed Motor
Dreame began as “Sky Workshop,” a Tsinghua student entrepreneurship group. Its founder, 俞浩, studied aerospace engineering; 郭人杰 calls him the inventor of China’s first tri-rotor drone. Many of Dreame’s R&D leaders today were junior or senior students in Sky Workshop at the time.
Around 2015, the team set 2 conditions for entering a new industry: the business had to be driven by technology rather than price, brand or marketing, and it had to serve millions of households by “improving the efficiency and happiness of every family, and ultimately raising the productivity of society as a whole.” Cleaning appliances became the first commercial application for its aerospace technology.
When international brands were operating at roughly 100,000 rpm and domestic suppliers at only 20,000-30,000 rpm, Dreame used a different technical path to reach 125,000 rpm in about 1.5 years. After its 2017 breakthrough, it incorporated and entered the Xiaomi ecosystem; 郭人杰 says Dreame subsequently became the world’s first developer to achieve a 150,000-rpm high-speed motor.
The 150,000-rpm milestone created more than an R&D challenge. Existing international components could no longer be reused, so the team rebuilt a domestic supply chain centered on Suzhou and extending roughly 100 km. And because no factory could manufacture the motor, Dreame had to build its own production line. The line went from roughly 50 operators to a lights-out factory; at the time of recording, it was mass-producing 180,000-rpm motors and had 200,000 rpm in reserve.
2. “Heart Plus Brain” Turned a Single Technology into a Robot Product Line
The high-speed motor first extended into vacuum cleaners and hair dryers; Dreame entered robot vacuums only after its algorithms had matured. 郭人杰’s analogy is that the motor provides power and suction and serves as the robot’s “heart,” while mapping speed, obstacle avoidance and other algorithms form its “brain.” Only “heart plus brain” makes a complete product.
When Dreame entered robot vacuums, Roborock, Narwal and Ecovacs had already spent years in the category, while iRobot still held significant brand strength and market share globally, especially in Europe and the US. The opportunity was not an empty market, but the fact that Chinese products had begun to lead even as Chinese brands had yet to fully match them.
Dreame’s industry view breaks manufacturing into 3 links: R&D, manufacturing and brand. The previous generation was broadly “R&D in the US, manufacturing in China, brand in Europe.” In the next generation, the quantity and quality of Chinese engineers could bring R&D back to China quickly, while the supply chain was already there. Suzhou’s role in producing roughly 70% of the world’s cleaning appliances reinforced that view.
The direct implication was clear: if Dreame could make R&D and the supply chain No. 1, then make the brand No. 1, the 3 links could converge and produce a world-leading company. That was the underlying reason it insisted on building its own brand even after it already had orders from the Xiaomi ecosystem.
3. Capital Opened the Window; Xiaomi and Early Movers Added Capability and Confidence
卫诗婕 noted that hard-tech startups enjoyed a capital tailwind after 2017, making it easy for many companies to raise money and pursue short-cycle expansion. 郭人杰 acknowledged that Dreame was not the fastest mover, but said the difference was sequencing: “R&D, supply chain, marketing, brand—not marketing, supply chain, R&D.”
He described the path of a number of new-consumer brands as raising money, making products, spending heavily on channels and marketing, then going back to fill in R&D and supply-chain gaps. Dreame instead used the Xiaomi ecosystem to refine vacuum cleaners, early robot vacuums and hair dryers, while learning about Chinese consumers and parts of the global market. Marketing was an amplifier of the product, not a substitute for it.
Capital also explains why few OEMs with solid technology had previously dared to build their own brands: a proprietary brand requires years of upfront investment. More important was the example set by Huawei and DJI, which crossed the “unexplored territory” first, followed by Anker, EcoFlow and Insta360, proving that Chinese hardware could build global brand influence.
郭人杰 says the previous generation of Chinese companies going overseas relied mainly on manufacturing and value for money. This generation had already built an R&D advantage, so the natural question became: “Why should the best product in the world be sold at a lower price?” Technology, capital and role models all had to arrive before building a proprietary brand became an executable option rather than a slogan.
4. High Prices Are a Financial Prerequisite for the R&D Loop, Not Brand Decoration
Dreame’s internal loop is straightforward: invest more in R&D, build better technology and products, earn higher profits, then fund the next round of R&D. If the best product can generate only low profits, the loop breaks; from the consumer’s perspective, “better yet cheaper” also creates a recognition problem.
In its early overseas expansion, the team first relied on product reviews to prove performance. In the second phase, it realized that the low-price market already had powerful Chinese incumbents, while the high end offered room. 郭人杰 says Dreame tried to enter and win China, Germany, France and Italy at the highest price points rather than using value for money as its wedge.
Dyson was an early reference point. It used a single breakthrough in high-speed motors to reshape vacuum cleaners and hair dryers and generate substantial profits. Dreame acknowledges that Dyson’s product experience was an industry benchmark at the time, but believes innovation beyond the motor was limited, so it made “surpassing Dyson” concrete through motor technology and product experience rather than a brand slogan.
One vacuum cleaner was deliberately priced above Dyson, yet its return rate became the lowest in Dreame’s history. The research explanation was counterintuitive: these consumers simply wanted “a vacuum cleaner better than Dyson” and were not price-sensitive; they might not even use it personally. With higher time costs, they were also less likely to return the product repeatedly over minor details.
5. The 2020 Breakout Was External Tailwinds Colliding with Years of Internal Preparation
Dreame did not predict the pandemic, but it had already prepared its products, supply chain and market understanding. Dollar liquidity and the capital boom brought by the pandemic, combined with more time spent at home and stronger demand for cleaning, opened the window. In 2020, the team went overseas first to test the global potential of premium pricing and product innovation.
After achieving an initial overseas breakthrough, the China business began pushing in earnest around the 2021 618 shopping festival. It went from zero to roughly RMB300M that year and grew to RMB3B in 2022; 郭人杰 says the China target for the year of recording was around RMB6B, while the program’s opening cited a company-wide figure close to or at RMB10B.
The scale curve was steep, but 郭人杰 rejected the idea that it came from “earth-shattering, astonishing marketing ability.” His explanation was that the 2017-2019 Xiaomi ecosystem phase had already refined multiple product lines, so Dreame arrived with technology, products and supply-chain reserves that competitors lacked.
That preparation translated directly into a time advantage. Where a peer might need 6 months to launch a product, Dreame could do it in “3.5 months.” Shorter iteration cycles allowed it to capture more windows, launch more products and keep pushing the industry through successive upgrades. The traffic dividend helped, but “technology and product are still our underlying color.”
6. Dreame Is Not Just Using China’s Supply Chain; It Is Raising Its Technical Ceiling
郭人杰 uses the high-speed motor to illustrate the two-way relationship between a brand and its upstream suppliers. Without Dreame’s demand, China’s relevant supply chain might still be stuck at 20,000-30,000 rpm. Once the company pushed its technical reserve to 200,000 rpm, suppliers acquired the capability to produce at a roughly 10x higher level.
The upgrade means the supply chain is no longer trapped at the low end of the “smile curve”; its value added should continue to rise and attract overseas orders. Average selling prices for robot vacuums also rose from RMB1,000-2,000 when Dreame entered the market to roughly RMB6,000 for Dreame products. 郭人杰 stresses that this was the result of joint iteration by multiple domestic manufacturers.
The risks are equally clear. Geopolitical shifts and US-China relations are pushing companies to build overseas factories and rebalance their supply chains. At the same time, scaled brands are moving from an OEM-heavy model toward owned factories to gain more control over cost, capacity and delivery stability. Dreame’s own motor line was an early example of that path.
7. Dreame Splits Its 12% R&D Investment into Defending Today and Betting on Tomorrow
“Effective innovation” starts with a change in perspective. In a To B business, the company is close to the customer but far from the end user; products are defined by the customer, with quality, volume and cost as the priorities. In To C, those internal metrics cannot substitute for user experience. The consumer’s pain point has to become “the starting point of our product.”
Dreame says it invests roughly 12% of its annual cost base in R&D, split into 4 buckets of 3% each. The first covers all existing technologies in the industry to avoid gaps; the second pushes each performance metric further—for example, increasing suction from 10,000 to 12,000 and reducing noise from 60 to 56; the third brings technologies such as autonomous driving into cleaning; and the last builds future categories such as robotic lawn mowers.
This is also 郭人杰’s response to claims that Dreame resembles a capital game and lacks substance: “Everyone has mistaken this 3% for our 12%.” His point is that the first 9% still covers the current industry, performance improvements and innovation; the cross-industry ambitions outsiders see in presentations account for only the final 3%.
The corresponding “4-dimensional R&D method” is to benchmark and surpass the industry’s best, expand into new functions, pursue cross-industry innovation and disrupt categories. Dreame acknowledges that entering a new category may initially require a relatively similar product to get into the market, but says the long-term goal must be innovation that consumers can perceive and are willing to pay for.
8. Not Chasing No. 1 in a Single Category Is Dreame’s Deliberate Trade-Off against Price Wars
卫诗婕 put the challenge directly: if Dreame makes such strong hair dryers and robot vacuums, why is it not No. 1 in domestic market share? 郭人杰’s figures were that wet-dry floor cleaners had reached No. 2 by the end of 2021, with share rising from the low teens to the low 20s in 2022; robot-vacuum share rose from roughly 5% to around 20%, but the leading competitor was still about 1.5x its size.
Dreame holds roughly 60% of the high-end robot-vacuum market but only about 20% across all price points because the high-end segment is smaller. To take the overall No. 1 spot, it would need to fill out the low-end price bands and increase its marketing budget—effectively launching a price war and competing for existing brand mindshare.
The team ultimately “hit the brakes.” Once a price war begins, it is difficult to stop, and it could damage the business model of both Dreame and the broader industry. 郭人杰 divides technology’s mass-market reach into 2 roles: “Some companies do technology, while others do mass affordability.” Dreame chose to keep building the best technology and leave cost reduction and mass adoption to others.
He still left room for offense: if Dreame decided to cut prices and spend aggressively, it “could have a chance to become No. 1.” 卫诗婕 then reframed No. 1 across categories: Dreame says it is the only brand to have built both robot vacuums and wet-dry floor cleaners, making it the No. 1 brand in the cleaning-appliance industry on an overall basis.
9. “Vacuuming Instant Noodles” Compressed Complex Product Advantages into One Second of Visual Proof
In June 2021, Douyin e-commerce was shifting from white-label goods to brands, while home appliances remained a small category. Dreame saw an opportunity to take a shortcut as a late entrant. Wet-dry floor cleaners were also well suited to livestreaming: consumers already understood what a vacuum cleaner did, but one-pass cleaning of dry and wet waste required a powerful, intuitive demonstration.
The team tested spilled water, hair, cat food and milk before landing on “vacuuming instant noodles.” Noodles spilled on the floor are a household disaster everyone understands; handling them with a broom and mop is cumbersome and “undignified.” A wet-dry cleaner removing them in 1 second eliminated the need to explain suction, one-pass wet-and-dry cleaning or other specifications.
Once the scenario was selected, daily sales jumped from roughly RMB200,000 to RMB1M. A livestream might repeat the instant-noodle demonstration roughly 200 times in a day. 郭人杰 describes the method as a 3-way test: front-end content, livestream demonstration and target audience are combined simultaneously; several concepts can be tested in 1 day, with the highest-converting angle identified within 1 week.
The initial customer profile also overturned expectations. The team had planned to target “polished moms,” but urban middle-aged and older consumers converted more efficiently: they had only recently begun shopping on Douyin and could not distinguish Tineco from Dreame. Dreame therefore stopped fixating on a preset persona and focused instead on “which audience is more efficient.”
10. After the Traffic Dividend Faded, Marketing Competition Became a Battle between Algorithms and Organizational Efficiency
Once every brand started “vacuuming instant noodles,” the visual concept quickly became commoditized and traffic prices rose. 郭人杰 says: “At the beginning, when people were finding the dividend, efficiency was high; once everyone came in, efficiency fell infinitely.” The remaining advantages were continuous trial and error and organizational response speed.
He describes ByteDance’s algorithm as if it knows a product’s gross margin. If a brand initially has roughly 30 points of ROI headroom, the algorithm gradually pushes bids toward that level, bringing ROI to around 1:3. When the brand can no longer sustain it, the system loosens slightly; when a new high-margin product appears, it tightens again. The result is a constant struggle in which the platform keeps “figuring out the math.”
In response to claims that Dreame pursued an aggressive expansion strategy and even “lost RMB2B in a year,” 郭人杰 says the figure did not match the company’s revenue scale at the time. He acknowledged losses but said they were “really quite limited.” Dreame says it required a Douyin ROI of 1:10 internally last year, versus an industry average of roughly 1:2.5, using the extreme target to force the team to keep searching for better content and audience efficiency.
卫诗婕 invoked the “prisoner’s dilemma” in appliance marketing: when gross margins fall, brands still have to increase traffic spending to buy share, ultimately producing revenue growth without profit growth. 郭人杰’s answer is the product itself—innovations users understand and are willing to pay for create room to reduce marketing costs.
11. The X20’s Mechanical Arm Became a Lever for Higher Prices, Lower Costs and Growth
郭人杰 says Dreame achieved an unusual combination last year: higher product prices, lower marketing expenses and simultaneous gains in market share and scale. The conventional sales answer is usually “cut prices and spend more.” Dreame calls its result an “impossible triangle” and attributes it to a genuine shift from To B to To C.
User interviews showed that robot vacuums turn away roughly 5 cm from the wall to avoid collisions. Table legs, sofa bases and other complex environments make the dead zones even larger. 郭人杰 says some households ultimately leave more than 30% of their floor area uncleaned—a pain point long understood by existing users.
The team developed 37 edge-cleaning concepts, including extending the body, using a deformable irregular mop and “twisting its butt” to push the mop toward the wall. Consumers worried that the first would splash dirty water onto the wall and that the second would increase collisions. The chosen solution was an internal mechanical arm that detects the wall and extends roughly 5 cm.
The X20 launched at RMB5,299 when the industry’s average selling price was about RMB3,999. It broke out immediately around 618, entered Europe 2 months later and quickly became No. 1 in Germany. In livestreams, a full cleaning demonstration was unnecessary: simply turning the machine over to show the arm and mop made the value obvious. The innovation raised both the selling price and marketing efficiency.
12. The Penetration Breakthrough Was Not the 90% Watching from the Sidelines, but the 10% of Existing Users
The team initially accepted the industry consensus that robot-vacuum penetration was roughly 10%: ask the remaining 90% why they did not use one, solve those pain points and the market could grow 10x. Repeated research showed otherwise. Non-users could offer 10 or 20 reasons, and solving them one by one still did not guarantee a purchase.
The real resistance came from the 10% who had already used robot vacuums. Early standalone machines collided randomly and got stuck on thresholds, spreading the reputation that they were “not useful.” For an appliance costing at least RMB3,000, potential buyers naturally consult friends and online reviews. The 90% who do not buy may therefore reflect the spillover of a poor experience among the 10%.
Dreame shifted its approach: “It’s not about competing for that 90%; we should come back and compete for this 10%.” The first priority was to make existing users willing to buy a second machine and actively spread the word that the product worked. The next product previewed during the recording included a mechanical leg able to clear roughly 5-6 cm thresholds, targeting old problems such as robots being unable to enter balconies or leave bathrooms.
郭人杰 says domestic and overseas demand differences are currently smaller than expected. Once a product is validated in China and quality and delivery are stable, it can go overseas within 1-2 months. Shared trends include feature integration, freeing users’ hands and more granular needs among pet-owning households and people living alone in 30-50-square-meter homes.
13. Bilibili Proved That Young Users Will Pay—and Will Back Chinese Technology
Dreame encountered 2 objections when it entered Bilibili: young people had no money and could not afford a RMB3,000 appliance, while their obsession with anime meant they were “not doing serious work.” The young team’s rebuttal was simple: users who can spend RMB3,000 on collectible figures may also buy a robot vacuum.
More important was their formative environment. 郭人杰 believes Bilibili users are more accustomed to “looking at the world as equals” and believe Chinese technology can be the best. The same content posted on WeChat Channels often drew comments accusing Dreame of copying or exaggerating; on Bilibili, it was more likely to receive a triple tap of likes, coins and favorites, along with active support for domestic technology.
Dreame first used hard-tech stories, teardowns and demonstrations of underlying principles to move from zero to one, then gave up trying to control the narrative completely. Bilibili is highly segmented, and the team could not understand every community. It therefore gave creators only the selling points it wanted to convey and left the final narrative to people who truly understood the audience: “Trust the power of UP creators.”
The best example was a collaboration on W10. The creator used soy sauce to draw Along the River During the Qingming Festival on the floor, let it dry overnight and then had the robot mop it clean in a single pass. Viewers saw the cleaning performance while being drawn in by the creation itself. The video stayed on the trending chart for roughly 1 week and generated sales of 7,000-8,000 units.
14. Large Models Could Turn Robot Vacuums from Cleaning Tools into Mobile Household Terminals
Roughly 1.5 years ago, Dreame already supported basic commands such as “Hello Dreame, start cleaning.” After partnering with Baidu to integrate “DreameGPT,” the goal became to understand more natural, emotionally charged language—for example, “Clean this place properly”—and automatically translate it into multiple passes, higher moisture and stronger suction.
The longer-term value is not limited to semantic interaction. A speaker is a fixed terminal; someone in the bedroom may not be able to control a device in the living room. A robot vacuum, by contrast, is one of the few household terminals that moves continuously while combining a speaker, camera and two-way voice capability. It can therefore take on robot functions beyond cleaning.
Products with video and two-way voice can be used to check on elderly family members or pets, reducing the need to install separate cameras. The “one-click pet search” feature scheduled for formal release at the time of recording would have the robot actively search the home for a cat while allowing the owner to watch and communicate by voice in real time.
郭人杰 believes the robot could eventually become a “family butler,” but Dreame’s long-term vision is broader: the company wants to become a technology ecosystem, with robots representing only one part of it. The goal is to keep delivering the best technology across as many industries as possible rather than remain permanently confined to cleaning appliances.
15. Dreame Learns from the Best but Refuses to Copy Mature Processes Wholesale
In its early years, Dreame mechanically studied Huawei’s IPMS, IPD and “Five Looks, Three Decisions” frameworks, then quickly found that the processes slowed rapid iteration. The team instead studied Huawei across different stages: every product today is expected to be a major hit and a premium product, but the company also went through a “small steps, fast running” phase early on. Dreame wanted to learn what fit its current stage.
It studied premiumization and overseas expansion from Huawei, smart ecosystems from Xiaomi, appliance channels, distributor management and a combat-oriented organization from Midea, and blockbuster-product creation from Li Auto. 郭人杰 calls this “absorbing the strengths of many companies,” not searching for one company to replicate in full.
俞浩 is described as an exceptionally fast learner who deeply experiences products and is even “willing to sleep alongside these products,” while rapidly mastering marketing, sales, supply chain and overseas expansion. The team averages roughly 26-27 years old. Its youth keeps it close to users, but also makes “not believing in rules” an organizational habit.
郭人杰 recalls worrying that he had worked for only 1 year when 俞浩 replied: “I have never worked a single day,” and predicted that never having worked for someone else would ultimately be an advantage. Lacking established experience made it easier to admit mistakes: when a decision was wrong, there was no need for layers of argument. It was “just a matter of 2 seconds.”
16. The “More Formidable than Apple” Ambition Must Be Viewed alongside the Cost of a Young Organization
卫诗婕 observed that, after Apple, fewer brands believe they can “change the world.” 郭人杰 replied that this nearly touched Dreame’s internal target: “We think that in 5 to 10 years, we need to become a company more formidable than Apple.”
What Dreame most wants to replicate is Apple’s ability to commercialize technology successfully and turn one generation of leading products after another into brand faith. Its logic is that if it can be the best in one field today, it may eventually do so in 5 or more. If it continues to provide leading technology, it could also build its own pool of super-users and a brand system.
The ambition is supported by the belief among its young employees that “the world is a makeshift operation.” 郭人杰 says a 20-year lead often reflects resources arriving earlier; if a later entrant finds a more efficient path and works several times harder, it can catch up quickly. “Either don’t do it or be No. 1” was once mocked by visiting customers but still hangs inside the company.
He also openly disclosed the cost. An elimination mechanism that looks only at data and ignores educational background gives young people significant leverage, but damages the employer brand; people who are eliminated will naturally view the organization as lacking humanity. Dreame acknowledges the negative externality and says it understands those assessments.
The more concrete lesson came in 2021. The China business was then selling roughly RMB8M-10M per month and held only about RMB30M of inventory. Douyin suddenly reached RMB1M in daily sales, selling through the stock in 3 weeks, while hardware materials originally required 3 months to prepare. 卫诗婕 says the lead time was later compressed to roughly 2.5 months or 2 months from the original 3 months, but the breakout was still forced to pause until early 2022. 郭人杰’s conclusion: the drive of young people will inevitably “punch some holes”; the key is whether the organization can fill them quickly.