Summary

Livestream e-commerce has moved from a bonus period of scarce influencers and an influx of brands into a hybrid model led by in-store livestreaming and amplified influencer distribution.谢如栋 estimates that 80%-90% of mature brands’ revenue now comes from their own stores. A healthy mix would be 70%-80% in-store livestreaming and 20%-30% influencer-led distribution: influencers create reach and bursts of demand, while store livestreams absorb the A1-A5 audience asset. “Otherwise, you spend desperately on advertising outside, and after users come in, there’s nowhere for them to buy.”
遥望’s early edge was not livestreaming capability from day one, but using branded goods, repeat purchases and paid-traffic know-how to close the traffic gap.When it entered the market at the end of 2018, the team was “disadvantaged across the board,” and its second livestream fell from RMB10M in sales at the debut to RMB500K. 遥望 nevertheless refused to sell low-quality white-label goods and, beginning in May 2020, was the first to direct capital into Kuaishou’s advertising system. By bidding aggressively for volume and closing transactions at high per-order prices, it ultimately developed 于大公子 from an office worker into a主播 generating more than RMB100M per session—and more than RMB300M in November 2020.
The industry’s current profit squeeze is fundamentally driven by fear-induced assortment contraction and single-brand specials, not merely by traffic peaking.To hit GMV and KPI targets, operating teams increasingly sell only leading brands and hot products. That makes short-term numbers safer but sacrifices gross margin, repeat purchases and主播 stickiness over time. 谢如栋’s diagnosis is blunt: “He was supposed to open a shopping mall,” but has become a specialty-store salesperson; brands that overinvest in top influencers will also see their ROI continue to decline.
遥望 reached its Douyin influencer-distribution peak in 2022, but moved its attention to X27 too early and lost a year that could have been used to consolidate the core business.谢如栋 acknowledges that the move was intended to adapt to top-tier traffic spreading toward mid- and long-tail players, but says “we should have continued influencer distribution in 2023,” and that X27 should not have jumped straight from 20,000-30,000 square meters to 200,000-300,000 square meters. He calls it “a strategic decision error,” while maintaining that the project has not failed and may have better odds now that the industry is becoming more rational.
The other causal chain behind “more revenue, no more profit” is expansion outpacing the capacity of the company’s systems, institutions and management bench.X27’s mass hiring diluted the culture; the old management team could not support the enlarged organization, while newly hired middle and senior managers struggled to adapt to the company’s pace. GMV-first thinking also encouraged excessive paid traffic and giveaways. 谢如栋 has therefore stopped personally treating symptoms and is staking the company on an AI management system: “You can’t bet on any specific business.”
Co-created consumer brands are 遥望’s attempt to reconnect influencer-distribution capabilities, celebrity IP and supply-chain strength into a second growth curve.The sanitary-products brand 朵唯, developed with 黄子韬 and 吴跃, generated RMB125M in sales in its first 45 days and more than RMB400M cumulatively. Pricing has already been pushed to the limit, so “every day is Singles’ Day.” But subsequent brands such as 简签 show that the main bottleneck is not initial scale, but organizational coordination among 遥望, the new team and partners; the ability to replicate the model remains unproven.
Singles’ Day is shifting from a decisive breakout moment into an elongated peak selling season, forcing platforms and brands to return their operating focus to everyday business.谢如栋 found the year relatively flat, while 李翔 called it the “least noticeable year in eight years”: the cycle now starts in early October, consumers feel no urgency, and everyday prices are already low enough. The promotion is more like the gaokao: “You can still innovate in a mock exam, but you can’t innovate in the gaokao.” It is therefore unsuited to innovation experiments.
Livestream e-commerce is not in retreat; it is shifting from isolated blowouts by a few top主播 to more distributed growth across the broader market.谢如栋 believes the extreme effect of a single session selling RMB500M or RMB1B is no longer realistic, but the overall market is still growing. Hangzhou’s talent is moving among A, B and newly formed C and D companies rather than leaving the sector en masse. For 遥望, the key asset in the next phase is not more livestream rooms, but whether systems can compress middle- and back-office functions, retain front-line talent and make every small team “feel like a startup” again.

Deep dive

1. The Industry’s Evolution Ultimately Pushed Influencer Distribution Toward Store-Led Fulfillment

谢如栋 calls 2018-2019 the industry’s “naive period”:主播, brands and consumers all lacked an understanding of livestream e-commerce, and the product mix was dominated by low-quality white-label goods. White-label products offered high margins but little stickiness; 遥望 saw its opportunity in bringing branded goods into the channel.
By 2020, leading brands were entering faster than the supply of主播. Quality products may have accounted for only about 10% of the market and were captured by a small number of主播, while supply exceeded demand for the rest. The livestream business also became profitable that year, entering a breakout phase of “making big money” in the second half.
2021 was the year Douyin e-commerce genuinely began to scale, while platform competition reached fever pitch in 2022 and nearly all brands completed their livestream-e-commerce education. Brands could not yet build teams or recruit talent themselves, leaving influencer-distribution agencies in control of scarce sales capabilities.
In 2023-2024, brands shifted toward a hybrid of store livestreaming and influencer distribution. Because costs in Hangzhou were too high, some store-livestream teams returned to the cities where their brands were based, but the most important livestream rooms remained concentrated in Hangzhou and Guangzhou. By the second half of 2024 and into 2025, 80%-90% of revenue for well-run brands was trending toward store livestreaming.

2. 王祖蓝 Was Not Icing on the Cake—He Was 遥望’s First Flag for Winning Branded Goods

A matchmaking experiment during Singles’ Day 2018 taught 谢如栋 that “follower count is useless”: two accounts with 10M followers each could generate completely different commercial output. The “Red-Hot Influencers, Good Goods” matchmaking business lasted only a month or two, but it exposed the difficulty of identifying主播.
Among the top 10 Kuaishou accounts by followers at the time, 王祖蓝 was the only entertainer. The other leading internet celebrities already had traffic, brand-acquisition and livestreaming capabilities, so “we were completely useless to them.” Entertainers crossing into a new industry, by contrast, needed institutional support. 王祖蓝 had also just come off 《奔跑吧兄弟》 and was at a peak in popularity, making him 遥望’s natural partner.
李翔 asked whether another entertainer could have succeeded. 谢如栋 refused to mythologize the outcome as a talent for spotting people: at the time, he had no idea whether 王祖蓝 was suited to livestreaming—“I hadn’t even met a celebrity before.” What he did know was that without this flag, 遥望 would have had “not even any goods,” let alone the ability to incubate ordinary主播.

3. Falling from RMB10M to RMB500K Confirmed That the Industry Could Be Learned

王祖蓝’s debut in Hong Kong generated about RMB10M in sales, with five or six of the top 10主播 present, several other celebrities involved and the novelty of a first show. A month later, the second session in Hangzhou sold only about RMB500K. 遥望 had assumed livestreaming would work like a shelf—“RMB10M on day one, RMB11M on day two”—but the novelty disappeared and the team did not know how to broadcast or operate.
谢如栋’s final explanation for the failure was one sentence: “The core reason was simply that we were too bad at it.” He asked other主播 for advice, and they bluntly replied, “You don’t know how to do any of it.” That answer actually ended his confusion: something that could be learned was not fatal. If the team and entertainer were already good but still could not sell, then the track itself might not work.
李翔 asked whether the company, entertainer and platform had reacted sharply. 谢如栋 said there was “not much reaction”; they thought for two or three days and continued. Giving up after “only two shows” would have amounted to sentencing the business to death, so the early losses were treated as normal tuition.

4. Brand Repeat Purchases and Ad Bidding Together Created 于大公子 in the Kuaishou Era

谢如栋 reduces 遥望’s early first principle to one idea: “A firm commitment to repeat purchases is the core competitive advantage.” 于大公子 primarily sold branded goods rather than the low-quality white-label products that offered higher margins at the time. The first transaction could lose money; as long as the product was good enough, the second purchase could be profitable.
After Kuaishou launched its advertising system, most主播 believed they had organic traffic and did not need to pay for it. 遥望 lacked traffic and therefore saw system-based bidding as the fairest mechanism. 谢如栋 says the system was called 磁力金牛, originally 小电通; beginning in May 2020, the company stopped paying for PKs and co-streaming, committing instead to “branded products plus system-driven traffic.”
His advertising experience created a pricing advantage before competition emerged. If the second-place bidder offered RMB5, 遥望 would set its ceiling at RMB20, while the actual payment might, in his understanding, be only RMB5.10. A competitor trying to overtake the top bid would have to go directly above RMB20. “I would generally set the price at the maximum,” causing large volumes to flow to 遥望.
The method turned an office worker into a主播 generating more than RMB100M per session within a year. In November 2020, 于大公子 sold more than RMB300M in a single session, ranking second on Kuaishou behind 辛巴. He delivered roughly 10 livestreams above RMB100M in 2021, which 谢如栋 described as “the result of the entire company giving everything it had.”

5. Douyin’s Lead Was Not a New Talent; It Was Two Years of Kuaishou Experience Transferred

By the time 遥望 entered Douyin, it had already spent two years on Kuaishou, and entertainers naturally carried across platforms. Against new agencies that did not even know where to click the “yellow shopping cart,” the company held a clear advantage in personnel experience and operating SOPs. 谢如栋 describes it as having “read two more years of books and worked two more years.”
2022 was 遥望’s strongest year in Douyin influencer distribution. A large number of entertainers, including 张柏芝 and 贾乃亮, joined. On one day, four to six livestream rooms had nearly 2M viewers online in aggregate, and 遥望 also represented a large share of Douyin e-commerce’s online scale at the time.
The first 遥望主播 to break RMB100M in a single Douyin session was “probably 贾乃亮,” 谢如栋 recalls. Few主播 on the platform had reached that level—perhaps one or two—but he also says he cannot remember the details and may be wrong. 罗永浩’s debut is the precedent he can recall with certainty. The lead here refers specifically to influencer distribution, not brand-owned store livestreaming.

6. Store Livestreaming Handles Everyday Conversion; Influencer Distribution Elevates Brand Assets

谢如栋 expects leading brands to settle into a core structure of “top celebrity tie-ups plus store livestreaming,” rather than relying on pure influencer distribution. Influencer distribution provides influence, endorsement and bursts of demand, but sustained conversion ultimately has to return to the brand’s own store.
Under the A1-A5 asset framework he uses, pure influencer distribution creates extensive reach without a stable setting to capture people at the top of the funnel. Once the主播 goes offline, users lose their point of purchase. Store livestreaming is therefore not merely a substitute channel; it is the infrastructure that prevents spending on external seeding, KOCs, KOLs and influencer distribution from being wasted.
谢如栋’s healthy mix is 70%-80% store livestreaming and 20%-30% influencer distribution: “A large amount of daily sales should happen in your own store,” while influencer distribution expands reach. He compares pure influencer distribution to advertising without opening a shop; pure store livestreaming, meanwhile, may never get off the ground because it lacks an initial audience asset.
遥望 has designed three business lines accordingly: continue providing influencer distribution for brands; co-create brands with supply-chain partners and celebrities; and bring the store livestreaming and celebrity clip accounts of co-created brands back into its operating system. The goal is for the influence generated by a breakout to keep producing revenue after the主播 stops streaming—a so-called “sun that never sets.”

7. Selling Only Leading Brands Makes GMV Safer While Compressing Gross Margin and Stickiness

谢如栋 acknowledges that 遥望’s assortment logic in 2023-2024 had drifted away from its early advantage. Product-selection staff once accounted for half of the operating team; now they may represent only one-tenth. The team repeatedly sells a small number of leading brands, avoiding the need to conduct real brand acquisition or study new categories, making short-term sales targets easier to hit.
The first cost is gross margin. Leading brands already have their own traffic and bargaining power: “They can’t possibly give you that much money—they need to make money too.” The more heavily an agency relies on easy-to-sell products, the more likely it is to generate revenue without profit.
In the past, 遥望 would recombine a brand’s A, B and C products with giveaways on the spot, recalculate the price and sometimes force the brand to call its accountant to confirm costs. Now brands simply provide three mature packages—A, B and C—“with no D.” Agencies have little negotiating room and no ability to create differentiated assortments.
When multiple influencers sell nearly identical packages, a consumer who does not buy in one livestream may see the same offer next door and conclude that it is “not scarce—forget it.” 谢如栋 contrasts this with 与辉同行: its stickiness comes from having different products every day, while an influencer should be a “recommendation hall,” not another brand-owned store livestream.

8. The Real Driver of Single-Brand Specials Is Not the Platform but Operators’ Fear

李翔 tried to identify the external force behind assortment contraction. 谢如栋 repeatedly answered: “Fear.” If a mixed session sells RMB5M, while a special featuring a hot product sells RMB20M, an operator carrying a KPI will inevitably choose the safer option that requires less preparation.
He compares livestreaming to diving. An athlete may train for 8-10 years and still feel fear on the platform, so they avoid high-difficulty moves. Mixed sessions and new products carry high failure rates; specials require familiarity with only one or two products, while paid traffic is more efficient. The entire industry therefore began collectively “playing it safe” from the second half of 2023.
When 谢如栋 managed operations personally, he prohibited frequent specials and allowed each主播 no more than one per month. Once management was delegated, the short-term numbers from specials looked strong, and both internal teams and peers began copying the model. He admits that 遥望 may have “led the entire industry astray,” and now wants to restore diversity while accepting that the adjustment will inevitably be painful.

9. Influencers Have Degenerated from Shopping Malls into Specialty Stores, Taking Brand ROI Down with Them

What concerns 谢如栋 most is not selling less in a single session, but losing the consumer mindset of browsing. A diverse assortment makes users unsure what they will encounter today and willing to return to see; repeated specials turn the主播 into “a specialty-store salesperson, when he was supposed to open a shopping mall.”
Brands have fallen into the same trap. Many over-rely on top influencers to run large specials, achieving sufficient user penetration without building stickiness and audience assets in their own stores. As the same users are repeatedly reached, influencer-distribution ROI declines, creating “a misconception on both sides.”
Pit fees have consequently evolved into “special-session fees”—essentially a large slot fee for one brand to occupy the entire session. 谢如栋 wants to return to a model with low pit fees, large mixed sessions and commissions as the primary source of earnings, allowing new brands to test at low cost. A middle ground with two or three brands rarely gets used because traffic ownership is hard to assess and neither side wants to pay.

10. 厂代, 艺统 and 遥望云 Broke a Nonexistent Industry into Specialized Roles

In the early days, many brands simply handed products to agencies without knowing how to turn selling points into livestream language. 遥望 was the first to establish a 厂代 department, staffed largely by people from television shopping. They represented manufacturers in organizing product logic and designing scripts, while also serving as assistant主播s who could explain products clearly.
谢如栋 believes the department was enormously helpful to 遥望, and that many of the industry’s later outstanding 厂代 came out of the company. Television shopping operated on a different platform, but had accumulated the scarce ability to “describe a product in full,” creating transferable knowledge for the early livestream-e-commerce industry.
遥望 also established an 艺统 department to act as a translation layer between entertainers and operators. The two sides had different ways of thinking and speaking, and direct communication had a “very large gap.” 艺统 served the主播 on one side while translating operating requirements into language entertainers could accept, preventing conflict from consuming execution efficiency.
Roles such as floor director, job boundaries and livestream SOPs gradually took shape through practice. Early group leaders handled both directing and floor control; only later were the roles separated. 遥望云 began development in 2020, codifying product selection, hot products and workflows, but became so complex that 谢如栋 says he “wouldn’t know how to use it today.” It has become a direct target for AI overhaul.

11. Traceability Livestreams and Flash Sales Show That Many “Formats” Began as On-the-Fly Fixes

谢如栋 remains cautious about traceability livestreams: 遥望 may have been among the earliest on Douyin, but he cannot say the same for Kuaishou. 于大公子’s first 韩束 special was held at the brand’s factory and generated about RMB5.2M in sales. The case made the factory setting part of both product credibility and the content itself.
“Flash sales” were more like an invention under pressure. 于大公子 had strong traffic, but the company did not have enough products already entered into inventory. The team therefore spread unselected products priced below RMB30 across the warehouse floor and cycled through goods priced at RMB9.90-RMB29.90 every 10 or 20 seconds. Products sold out before they had even been introduced.
谢如栋 summarizes the mechanism: low prices reduce decision costs, rapid product turnover keeps users waiting for the next item, and per-minute conversion is useful to the platform’s traffic algorithm. A two-hour session could sell about RMB2M. The format was later reused on Douyin, where peers copied it.
These innovations were not permanent formulas. After Douyin’s traffic mechanism changed, flash sales remained usable but were no longer as effective as in the early period. 谢如栋 uses the example to warn that platform tools and algorithmic windows change; a tactic from a bonus period should not be mistaken for a durable capability.

12. Entertainers Provided the Commercial Base; Ordinary主播s Provided Low-Cost Training

遥望 began developing ordinary主播s such as 于大公子 about four months after signing 王祖蓝, in August or September 2019. 于大 initially served as 王祖蓝’s assistant主播. 谢如栋’s logic was to let the entertainer establish the brand-acquisition and branded-goods base first, then allow ordinary主播s to stream frequently and make mistakes.
遥望 did not continue signing entertainers in rapid succession. Its second entertainer, 王耀庆, did not join until 2020. “The first one hasn’t even worked yet, and you don’t know how to do it—so why sign so many?” Repeated mistakes by entertainers could damage partners’ confidence; ordinary主播s were cheaper to train and better suited to refining the team.
The industry broadly experimented with incubating主播s, but many lacked brand-acquisition capability and could sell only apparel and other non-standardized goods, resulting in low success rates. 遥望 took a dual-track approach of “entertainers plus ordinary主播s.” 王祖蓝 carried the flag, while 于大 and others proved that an agency could cultivate sales capabilities.

13. Replicating the Model to the Tenth or Fifteenth 主播 Exposed the System’s Efficiency Decay

谢如栋 does not believe top主播s can be replicated indefinitely in their original form. The first teams had long practical experience and received the entire company’s attention and resources. By the tenth or fifteenth IP, personnel had changed, attention was spread thinner and team quality naturally declined.
遥望’s data was still growing in 2023, but its operating precision had already deteriorated. Once GMV became the core metric, teams leaned toward buying more traffic and giving away more products, investing money that should not have been spent into scale. Profit consequently failed to keep pace with revenue.
His review does not demand that every new employee be naturally excellent. It emphasizes “the excellence of institutions and the excellence of systems”: institutions handle rewards and penalties, while systems determine operating logic. If both are weak, newcomers receive no training and cannot tell whether they are learning or merely coasting. Expansion then amplifies organizational noise.

14. X27’s Direction Came from a Correct Forecast; the Mistakes Were Starting a Year Early and Scaling Too Fast

In 2022, 遥望 judged that livestream e-commerce would shift from concentration among a few top players toward growth among mid- and long-tail participants. It therefore planned X27 to serve more主播s and agencies with more accessible pricing, high-end offline settings and operating capabilities, while countering the industry cycle.
谢如栋 now acknowledges that “we should have continued influencer distribution in 2023.” Pulling resources out early to build X27 was a strategic timing error. Influencer distribution had only just stabilized in 2022; replication efficiency was declining, but there was still room to grow. He should have launched X27 in 2024 rather than shifting the center of gravity in 2023.
The scale decision was also too aggressive. The project’s building area exceeds 250,000 square meters. He believes the more rational path would have been to start with 20,000-30,000 square meters, accumulate commercial and management experience, then expand to 200,000-300,000 square meters. “We moved too fast.”
He continues to distinguish between a failed decision and a failed project. The direction has not been invalidated, and the project is still moving forward. As the industry shifts from frenzy to rationality, speculative expectations among mid- and long-tail agencies have weakened, potentially making them more receptive to stable enablement. X27’s window may now be more favorable than it was two years ago.

15. X27 Strained Not Just Rent Economics but Cultural and Management Capacity

Starting a year early first meant higher property rents. Second, the team had not yet been built. Third, the industry was still in full swing, salaries were inflated and job-hopping was frequent. 遥望 was itself too busy to fully deliver the operating enablement promised to tenants.
谢如栋 estimates that industry salaries have now fallen to roughly 70% of their peak. At the same time, workers who were once one-year beginners have become three-year “veterans,” and their capabilities have improved. He believes current compensation is more aligned with industry reality than the earlier level.
A large project brings in many new people at once, diluting the existing culture. The old management team could not support the additional organization, while externally hired middle and senior managers struggled to adapt to 遥望’s pace, causing many to leave. 谢如栋 sees these three factors as the most serious organizational problems created by rapid expansion.

16. “More Revenue, No More Profit” Reflects Falling Industry Gross Margins Plus Company Execution Detours

Facing criticism after the third-quarter financial results, 谢如栋 acknowledged that the outside assessment was reasonable. In 2024-2025, the impact and execution problems at X27 coincided with declining gross margins in influencer distribution, creating two simultaneous pressures and magnifying the “more revenue, no more profit” outcome.
He maintains that the direction was correct but does not avoid the management shortcomings: “The problems over these two years were still fairly serious.” Too many innovation businesses were launched before the organization was ready; communication and understanding up and down the chain were slow, ultimately disconnecting strategy from front-line execution.
The adjustment therefore is not to keep adding business ideas. 谢如栋 says 遥望 “doesn’t lack businesses, and it doesn’t lack business ideas.” Its real shortcoming is the execution capacity of its young teams. Until that is repaired, further expansion will only make the organization lose control faster.

17. The AI System Has Been Put First Because 谢如栋 Has Given Up Personally Treating Symptoms

谢如栋 estimates that returning to the livestream floor himself might generate only about 20% growth for the year. With too many livestream rooms, however, the long-term value would be limited. “I’m not treating symptoms anymore.” He has not attended livestream-team meetings for a long time, focusing his limited energy on system management.
遥望 originally planned to develop its AI system in parallel with business expansion. Development began in January 2023, with the system expected to be usable in March or April, but the company hired ahead of schedule while the system was still being tested. 谢如栋 now believes the sequence was wrong: “System construction cannot happen in parallel; it has to be completed first.”
The system’s robustness has improved markedly over the past two months. It used to “crash as soon as it ran”; now it generally does not, meeting the basic conditions for deployment. 谢如栋 still describes this as the final difficult stretch: “If we grit our teeth, it may come out.”
His bet comes with a clear awareness of the risk. If the system succeeds, 遥望 can replicate knowledge, processes and staffing configurations across more livestream rooms and partner brands. If the company continues relying on people to monitor people, the business will only deteriorate as it expands. “You can only bet on this. You can’t bet on any business.”

18. From Games to Livestreaming, 谢如栋 Repeatedly Admits That Exiting the Second Curve May Also Have Been a Mistake

In 2017-2018, 谢如栋 devoted most of his attention to capitalization, communicating with dozens of listed companies, regulators and brokerages. The public-account matrix business was relatively simple and led by other managers. Only after the listing did he shift his focus back to livestream e-commerce.
The games business lasted five or six years. Monthly channel distribution reached RMB100M-RMB200M, close to the ceiling. As licensing tightened, mobile-market channels declined and new games became scarcer, profits grew too small and 遥望 chose to exit. 李翔 noted that peers who stayed in the market later saw monthly top-ups return to RMB50M or even RMB100M. 谢如栋 responded bluntly: “Actually, we should have persisted too.”
遥望 also tried developing games in-house, but 谢如栋 compares it to making a film: when a blockbuster will appear is uncontrollable. Mature companies often have a foundation game that funds new teams; 遥望 had neither that asset nor the inheritance gene, making the risk too difficult to absorb.

19. He Does Not Believe in a Fixed Circle of Competence, Only in Whether an Organization Can Decide to Learn

李翔 asked whether 遥望 was inherently bad at in-house development, content and products. 谢如栋 pushed back: when the company moved into livestreaming, the original team was equally “unsuited,” but it could be reshaped once the decision was made. If personality can change, so can organizational structure.
谢如栋 describes himself as somewhat socially anxious. At a dinner with 9 people where he knew only one, he might leave without adding the person beside him on WeChat. If the dinner involved an important deal, he would first spend 5 minutes mentally preparing and force himself to speak. He also overcame blushing on stage through repeated practice.
This does not mean he likes selling goods. He would rather build products and systems, and dislikes charging directly or haggling. But when business requires it, he still learns. “The teacher may not necessarily know how to do it himself.” He sees himself as a director or coach: he may not be able to become a good主播, but he can design the product mix, pricing and livestream method.

20. Kuaishou Taught Him That Emotional Value Could Be Converted, Not Just Product Value

Public-account open rates had not yet fully collapsed, but 谢如栋 predicted that homogenized content would decline. Game-distribution scale was also limited, and the listed company needed a third, larger track. After watching 散打哥 generate more than RMB100M in a single livestream, he felt for the first time that the market was large enough.
Taobao Live had not produced the same conviction.主播s carefully explained products, transactions were more conventional and the audience skewed more female, which he found “too bland.” Kuaishou livestreaming looked like a wholesale market hawking goods: products could sell before they had been fully explained, with content, bustle and emotion working simultaneously.
谢如栋’s conclusion was that “the consumer’s emotional value is higher than the value of his material purchase,” and that influencer distribution capable of instantly lifting the atmosphere was better suited to 遥望. When the company entered, the team had only just downloaded Kuaishou and neither buyers nor sellers knew how to make payments. But the listed-company identity, commercial logic, courage and execution were enough to support a small-scale experiment.

21. Singles’ Day Cooling, Co-Created Brands and Organizational Restructuring All Point to an Era of Everyday Operations

谢如栋 considers Singles’ Day in 2020 and 2021 the most satisfying: going from zero to one, breaking RMB100M repeatedly, with an atmosphere “like celebrating the New Year.” Now the cycle stretches from early October, consumers feel they can buy tomorrow or the day after, and everyday prices are already low enough. 李翔 called this year “the least noticeable one in eight years,” while 谢如栋 agreed that the year was “relatively flat.”
Promotions cannot carry innovation either: “How can you innovate in the gaokao? You can still innovate in a mock exam.” Singles’ Day can remain an excuse for a peak selling season, but it no longer determines the year’s fate. As the season stretches, it also creates a brief slowdown before Singles’ Day 2 extends the cycle, with limited practical meaning.
遥望’s co-created brands are operated directly at everyday low prices. 朵唯, developed with 黄子韬 and 吴跃, generated RMB125M in sales in its first 45 days and more than RMB400M cumulatively. Because gross margin has already been compressed sharply, 谢如栋 says “every day is Singles’ Day”; the promotional period creates some volatility, but not much.
For new brands such as 简签, sales are not the primary issue; team coordination is. Representatives from 遥望, newly hired managers and the brand side have been “assembled from all over,” and 谢如栋 believes it may take about 2 months to establish a working rhythm. He hopes that once the system goes live, partner brands can quickly manage through it, while celebrity IP, factory supply chains and 遥望’s marketing capabilities can be replicated across more brands.
Internally, the company also needs to return to smaller entrepreneurial units. The middle office must remain, but the back office cannot add staff dynamically with every new business. Otherwise processes slow down, shared costs rise and the money earned by the front line gets diluted. If the system can compress repetitive work in finance and other back-office functions, more earnings and autonomy can be returned to livestream teams.
Talent attrition is another systemic risk. 谢如栋 says many livestream companies’ core employees came from 遥望. Employees are often poached externally during the vulnerable phase when they are “about to be promoted but not yet promoted,” offered the post-promotion price like “a crab molting.” The company also once judged ability solely by absolute scale, underestimating front-line personnel whose IPs were small but whose capabilities were stronger.
On claims that influencers are leaving Hangzhou, he sees more of a structural transition. Talent is mainly moving between A and B companies while continuously giving rise to C and D companies. The concentration effect of a single top主播 selling RMB500M or RMB1B in one session may no longer be realistic, but the overall market is still growing. Hangzhou’s strength is explosive power and complete infrastructure; Guangzhou may be stronger on stability. What is being called a retreat is more accurately a shift from individual miracles to steady, group-wide growth.

The Chip That Could Unlock AGI.Kalshi CEO Tarek Mansour on Raising $1BN, CNN and CNBC Deals & the Polymarket Feud