WeChat: How Social Capital Becomes Economics
Social Capital as Variable Cost
In the family chat groups of WeChat, a distinct etiquette has emerged around the request that would otherwise constitute a minor social violation. A link arrives, “Join my team for Sichuan oranges, 19.9 RMB for five kilograms”, followed by a polite emoji, sometimes even an apology for the intrusion. The recipient clicks, joins the “team” of two or three or ten buyers, and secures a price roughly thirty percent below market rate. The sender has converted a modest portion of their social standing into hard currency savings, outsourcing to their personal network the customer acquisition costs that Alibaba and JD.com pay to search engines and banner ad networks. This transaction, repeated 788 million times annually, constitutes the foundational economics of Pinduoduo.
The mechanism appears simple: users unlock “team purchase” discounts by sharing product links with friends within a twenty-four-hour window. Yet this design externalizes the most expensive component of e-commerce, CAC, to the very users the platform serves. Where traditional marketplaces purchase traffic, Pinduoduo rents it, paying not in cash but in discounted goods. The result is a variable cost structure for acquisition that scales inversely with the density of a user’s social graph. A grandmother in a fourth-tier city with extensive family networks becomes more valuable than an isolated urban professional, flipping the demographic logic of premium e-commerce.
This commodification of guanxi, those reciprocal social obligations that traditionally lubricate Chinese commerce, depends on a fragile equilibrium. The discount must be large enough to justify the social friction, yet small enough that the platform retains margin. The twenty-four-hour window concentrates dispersed demand into discrete, predictable batches, transforming sporadic purchasing intent into manufacturing orders. But the model assumes that Chinese consumers will indefinitely tolerate the reputational labor of soliciting friends for marginal savings, that the embarrassment of spamming a chat group remains cheaper than paying full price. Should privacy norms shift, or should the social stigma of these requests harden, the primary distribution channel collapses. The business, in essence, mortgages the user’s social relationships against the promise of low prices, collecting interest in the form of aggregated demand.
The 24-Hour Manufacturing Window
The demand aggregation that begins in WeChat chat groups terminates in factory floors and agricultural cooperatives through a mechanism the company terms C2M, onsumer-to-manufacturer. Unlike traditional e-commerce, where orders arrive asynchronously and unpredictably, the team-purchase model concentrates dispersed demand into twenty-four-hour windows. A manufacturer receiving ten thousand orders for identical phone cases, all placed simultaneously, can optimize a single production run, bargain for upstream inputs at volume, and eliminate the waste of speculative inventory. This visibility into demand, rather than the absolute volume of orders, constitutes the platform’s value proposition to suppliers.
For agricultural producers, this architecture addresses a structural inefficiency that has long plagued Chinese farming. Online penetration for agricultural products remains in the mid-single digits, compared to thirty percent for electronics and apparel, largely because fresh produce cannot tolerate the inventory uncertainty of traditional retail. A farmer planting for anonymous demand risks crop failure; a farmer planting against confirmed team orders knows precisely how much to harvest and when. The twenty-four-hour window functions as a just-in-time procurement system for persimmons and garlic, matching lumpy supply with lumpy demand through the buffer of social coordination.
This aggregation creates a flywheel. Lower manufacturing costs enable lower prices, which make the social sharing incentive more compelling, which generates larger batch orders, which further reduce unit economics. The platform thus monetizes not merely the transaction between buyer and seller, but the temporal compression of demand itself. Traditional retailers charge for the convenience of immediate availability; Pinduoduo charges for the patience to wait for the team to form, capturing the value of deferred gratification and collective purchasing.
The extension of this logic into Duo Duo Grocery, launched in August 2020 and rolled out to three hundred cities within months, reveals the model’s logistical frontier. By utilizing local pickup points, convenience stores, laundromats, residential community centers, the platform avoids last-mile delivery costs that typically constitute one-third of total logistics expenses. Consumers collect their produce the day after ordering, accepting the friction of self-collection in exchange for prices that undercut supermarkets by forty percent. The model bifurcates the logistics network: long-tail exotic goods move through multi-day shipping, while perishables move through dense local fulfillment infrastructure. This split strains capital allocation, requiring simultaneous investment in national distribution for non-native products and hyperlocal density for fresh vegetables, yet the alternative, unreliable cold chain delivery, would breach the trust necessary for repeat purchases of perishable goods.
The Advertising Subsidy
Beneath the consumer-facing facade of team discounts lies a marketplace engine that generates revenue through a fundamentally different mechanism than the transaction fees that sustain competitors. Pinduoduo charges merchants a nominal 0.6 percent transaction service fee, described internally as a “rack rate” covering little more than payment processing. This near-zero friction creates a supply base of small agricultural producers and manufacturers who cannot afford the higher commission rates and mandatory advertising spend required by Alibaba’s Tmall or JD.com. The low fee functions as a recruitment tool, attracting the long-tail merchants who provide the exotic SKU depth, $3 electric toothbrushes, bulk rice from remote provinces, that distinguishes the platform.
Yet the company does not survive on half-percent commissions. The economic engine is advertising: search placement, in-feed promotion, live-streaming visibility. Merchants compete in auction-based systems for attention within a high-volume, low-friction marketplace. Because the barrier to listing is minimal, competition for visibility is maximal. A merchant selling socks may pay nothing to list, but must purchase advertising to appear above the thousand other sock sellers when a user searches. The platform thus extracts value not from the transaction itself, but from the merchant’s anxiety about being discovered.
This creates a structural tension between the purity of the marketplace model and the pressure for revenue growth. The 0.6 percent fee maintains the appearance of accessibility, preserving the ecosystem of small merchants who create the platform’s “treasure hunt” appeal. Yet the business depends on converting these merchants into advertisers. If merchants refuse to bid for placement, perhaps because margins on $2 phone cases cannot support customer acquisition costs, or if excessive ad load degrades the user experience, revenue growth stalls against the fixed infrastructure costs of maintaining a marketplace serving nearly eight hundred million users. The platform must balance the merchant’s ability to survive against the necessity of monetizing their desperation for visibility.
Colin Huang’s strategic discipline in maintaining this fee structure, even as the company scaled to exceed Alibaba’s user base, represents an economic variable rather than mere corporate culture. By resisting the temptation to raise take rates, the platform preserved the supply-side fragmentation that prevents any single merchant from gaining negotiating power against the algorithm. The decentralization of the merchant base ensures that advertising remains the only path to scale, locking in a revenue stream that scales with merchant anxiety rather than consumer spending alone.
The Entertainment Tax
The transaction is never purely transactional. Within the app, users water virtual trees, steal “droplets” from friends, and tend orchards that yield real fruit delivered to their door. Duo Duo Orchard and its companion games transform idle minutes into retained engagement, creating what the company terms “Disneyland” browsing behavior. A user intending to purchase toilet paper may spend twenty minutes harvesting digital fruit, viewing advertisements and merchant listings during moments that would otherwise belong to short-video platforms or mobile games.
This gamification solves a retention problem inherent in utilitarian commerce. If users opened the application only when they needed specific goods, the frequency of engagement would mirror necessity, erhaps monthly for apparel, weekly for groceries. By interweaving purchasing with entertainment, Pinduoduo captures daily active users who return to water their trees, incidentally browsing discounted merchandise during rituals of boredom. The agricultural poverty alleviation angle, orchard rewards sourced from farmers living below the poverty line, adds a layer of moral participation to the dopamine loop, allowing users to conceptualize their screen time as charitable labor.
Yet this design introduces a tension between entertainment value and transactional efficiency. The user seeking only to reorder rice finds their path obstructed by game mechanics, spinning wheels, and promotional pop-ups. Over-indexing on gamification alienates the utilitarian shopper who values time over savings; under-investing risks losing the idle browsers who generate advertising impressions. The platform must maintain a precise calibration where the friction of entertainment remains just tolerable enough that the price discounts compensate for the cognitive load. It is a retention tax paid in attention rather than currency, and like all taxes, it risks provoking revolt if raised too aggressively.
WeChat’s Shadow
For all its manufacturing innovations and gamified retention, the business rests on infrastructure it does not control. The viral acquisition mechanism and social graph intelligence depend entirely on Tencent’s WeChat, the super-app through which team links propagate. Pinduoduo exists as a mini-program within WeChat’s ecosystem, ingesting social data, who influences whom, whose recommendations convert to purchases, which family clusters buy similar produce, that creates a recommendation engine impossible to replicate on purely transactional platforms.
This dependency constitutes the defining structural constraint. Should Tencent restrict API access, demote PDD mini-programs in favor of its own commerce initiatives, or simply throttle the sharing mechanisms that enable team formation, the flywheel severs overnight. The proprietary social graph intelligence, knowing that User A strongly follows User B’s fresh produce recommendations, mapping influence patterns within familial networks, derives from WeChat’s infrastructure. Without it, the decentralized, trust-based merchandising collapses into the keyword bidding wars that characterize traditional e-commerce.
The platform has attempted to mitigate this through its standalone app, yet the data suggests that the majority of new user acquisition still originates in WeChat groups. The relationship with Tencent remains symbiotic but asymmetrical: Pinduoduo drives engagement within the WeChat ecosystem, yet lives at the mercy of platform policy shifts. This tether shapes every strategic decision, from the cautious expansion of first-party logistics to the aggressive pursuit of agricultural suppliers who might provide unique inventory unavailable on Tencent-affiliated platforms. The business is built on rented land, and the lease terms are subject to renewal by a landlord with competing commercial interests.
The Perishability of Trust
The model assumes that quality failures remain isolated incidents rather than systemic patterns, that a spoiled batch of apples does not poison the social graph that delivered the customer. When a user recommends a 19.9 RMB fruit box to their aunt, they collateralize their personal reputation against the platform’s fulfillment capabilities. If the produce arrives rotten, the loss extends beyond the refund; the recommender loses face, the social trust required for future team formations erodes, and the mechanism that aggregates demand begins to fracture.
This fragility intensifies in agricultural logistics, where the network lacks the cold-chain infrastructure necessary for consistent quality. During summer heat waves or demand surges, third-party logistics partners fail to maintain temperature controls for leafy greens, generating the precise category of complaint that spreads virally through the same WeChat groups that drive acquisition. Unlike electronics, where a counterfeit phone charger represents a financial loss, spoiled food triggers visceral disgust and safety concerns. The “low risk” tolerance of the initial transaction, the psychological comfort of spending only twenty yuan, converts to platform-wide reputation damage if the value proposition repeatedly delivers inedible goods.
The risk compounds because the platform’s merchants operate at margins that preclude generous return policies or quality guarantees. A farmer selling at 0.6 percent commission cannot absorb the cost of replacement shipments; the platform’s advertising-dependent revenue model provides no cushion for quality control investments. The result is a persistent tension between the price points necessary to fuel social sharing and the quality levels necessary to sustain it. The business lives at the edge of acceptable failure rates, optimizing for the threshold where disappointment remains insufficient to overcome the inertia of low prices.
First-Party as Admission of Failure
Despite the asset-light advertising model, Pinduoduo maintains first-party inventory operations, purchasing popular goods directly to ensure availability during viral demand spikes. This inventory functions as a stopgap, filling supply gaps when third-party merchants cannot scale to meet the lumpy demand generated by successful team-buying campaigns. Yet every unit of first-party sales represents a contradiction of the platform’s theoretical purity, introducing operational complexity, warehouse costs, and channel conflict with the third-party merchants who pay for advertising.
The tension is economic rather than merely operational. When the platform sells directly, it competes against its own advertisers, creating the perverse incentive to prioritize its inventory in search results over paying merchants. Expanding first-party operations risks transforming the business into a traditional retailer with thin margins and capital-intensive logistics; abandoning them risks stockouts during peak demand that break the viral loop. The presence of first-party inventory signals moments when the marketplace model has failed to align supply with demand, functioning as a strategic safety valve that, if relied upon too heavily, corrodes the advertising-based economics that justify the valuation.
This inventory represents the boundary of the platform’s ability to coordinate manufacturing through demand signals alone. When the twenty-four-hour aggregation window produces orders that exceed supplier capacity, the company must choose between disappointing users, breaching the trust that powers social sharing, or absorbing the capital intensity of physical goods. The optimal state is pure marketplace; the actual state requires increasingly sophisticated inventory buffers that threaten to drag the business into the low-margin logistics wars it was designed to avoid.
The model thus operates in a narrow band: sufficiently reliable to maintain social trust, sufficiently fragmented to prevent merchant concentration, sufficiently entertaining to drive daily engagement, and sufficiently cheap to justify the social labor of sharing. It is a machine for converting the small humiliations of asking friends for discounts into manufacturing efficiency, working only so long as the savings outweigh the shame, and only so long as WeChat allows the asking.
The agricultural supply chain assumption, that farmers can scale to meet viral, lumpy demand spikes, collapses upon contact with biological reality. A garlic farmer in Shandong province may receive, through the team-purchase interface, an order for twenty metric tons to be harvested, packed, and shipped within seventy-two hours, a volume that represents six months of traditional wholesale distribution compressed into a single manufacturing batch. The platform’s aggregation architecture treats agriculture as if it were semiconductor fabrication, assuming that capacity can flex upward given sufficient demand visibility. But garlic requires six months in the ground; it cannot be 3D-printed to meet a viral spike.
This temporal mismatch creates a bullwhip effect that travels backward through the supply chain. When a particular variety of mandarin orange achieves virality, perhaps because a KOL’s recommendation triggers ten thousand simultaneous team formations, the immediate demand far exceeds the ripening cycle of the fruit. The farmer, facing the platform’s rating algorithms that penalize stockouts with reduced visibility, has three options: harvest early and ship unripe produce, purchase complementary supply from secondary markets at higher cost and absorb the margin erosion, or default on the orders and face delisting. Most choose the first option, delivering hard green oranges that arrive inWeChat group complaints rather than repeat purchases. The platform’s data shows the transaction as complete; the social graph registers the failure as personal betrayal.
The perishability ceiling manifests most brutally during the climatic extremes that increasingly disrupt Chinese agriculture. During the summer of 2022, temperatures in Sichuan exceeded forty degrees Celsius for consecutive weeks, wilting the leafy greens that Duo Duo Grocery had promised for next-day pickup. The third-party logistics network, lacking dedicated cold-chain infrastructure, transported bok choy in unrefrigerated trucks through asphalt-melting heat. Customers arrived at their local convenience store pickup points to find bags of blackened, fermenting vegetation. The platform’s refund mechanism, automated and instantaneous, addressed the financial loss but could not repair the social damage. These were not anonymous Amazon deliveries; they were recommendations from mothers to daughters, from neighbors to neighbors. The failure mode was public and humiliating, exactly the emotional valence that breaks the sharing loop.
The assumption that agricultural supply can scale to meet viral demand ignores the working capital constraints of smallholders. A farmer who accepts a team-purchase order for ten thousand units must harvest, pack, and ship before receiving payment, floating the cash for labor and materials during the gap. Traditional wholesale markets pay upon delivery; Pinduoduo’s settlement cycles, while improving, still create a liquidity crunch for producers operating on margins of pennies per kilogram. The result is a selection bias toward larger agricultural aggregators who can finance the float, slowly transforming the long-tail supply base into concentrated middlemen who arbitrage between scattered farmers and the platform. This concentration erodes the very SKU diversity, exotic, non-native, hyper-local products, that distinguishes the marketplace from supermarkets. When the garlic farmer drops out due to working capital exhaustion, he is replaced by a trading company that sources generically, commoditizing the inventory that the algorithm promised was unique.
The Maturation of the Orchard
Duo Duo Orchard operates on a behavioral assumption that remains unproven beyond the novelty phase of platform adoption: that mature users will continue to water virtual trees for physical rewards after the gamification mechanics lose their dopaminergic punch. The current user base, which reached 788 million annual actives in fewer than six years, contains a significant cohort who adopted the platform during its hyper-growth phase, when the novelty of “stealing droplets” from friends carried the social currency of trend participation. As these users age demographically and the platform matures, the retention mechanism faces the test of sustained efficacy without novelty.
The game mechanics exploit specific cognitive vulnerabilities, variable reward schedules, loss aversion regarding withered trees, social obligation through “watering gangs”, that psychological research suggests habituate over time. A user who has tapped the screen ten thousand times to water a virtual orange tree may experience not satisfaction but resentment upon finally receiving the physical fruit, calculating the hourly wage of their labor against the street value of the produce. The “idle time” browsing that the platform monetizes through advertising becomes, for the maturing user, recognized as time waste. The recognition triggers either deletion or, more corrosively, silent abandonment: the user remains technically active but ceases the daily rituals that generate ad impressions.
The agricultural poverty alleviation narrative, orchard rewards sourced from farmers below the poverty line, adds a layer of moral complication to this fatigue. Initially, the user conceptualizes their screen time as charitable labor, extracting value from boredom to support rural development. Over time, the awareness that their engagement primarily enriches the platform’s advertising business, with poverty alleviation functioning as marketing veneer, may generate the specific cynicism that destroys trust-based platforms. The moral economy of the orchard requires that users never calculate the precise advertising revenue generated by their thirty minutes of daily play against the $3.50 wholesale value of the fruit they receive quarterly. Once that calculation spreads through the social graph, the entertainment value converts to perceived exploitation.
The platform’s response, escalating the complexity of games, introducing “Duo Duo Kingdom” and livestock breeding mechanics, risks accelerating the fatigue rather than alleviating it. Each new game layer adds cognitive load to a user base that originally joined for discounts, not entertainment. The tension between transactional efficiency and gamified retention resolves, for the aging user, in favor of the former: they migrate to the search bar, purchase the toilet paper, and exit without harvesting digital crops. The advertising impressions vanish, and the user reveals themselves as a low-lifetime-value customer acquired through unsustainable gaming subsidies.
The Throttling Scenario
The bear case for WeChat dependency is not theoretical abstraction but concrete scenario planning visible in Tencent’s peripheral maneuvers. The super-app’s parent has gradually elevated its own commerce initiatives, WeChat Pay’s merchant ecosystem, live-streaming shopping integrations, and internal search-based product discovery, while maintaining Pinduoduo’s mini-program access as a strategic hedge. The throttling, when it comes, will not arrive as a sudden API termination but as algorithmic demotion: Pinduoduo links will load three seconds slower than internal commerce links, team-purchase cards will render without preview images, and the “share” button will require an additional confirmation click that introduces sufficient friction to break the viral impulse.
Consider the specific mechanics of social sharing degradation. Currently, a user taps “share to WeChat,” selects a group, and the link generates a rich preview card displaying the product image, price, and team-formation progress. Tencent could degrade this to a plain-text URL, requiring the recipient to copy and paste into the Pinduoduo app, transforming a one-tap conversion into a four-step process. The twenty-four-hour team-formation window, already tight, would fail to close as users abandon the transaction midway through the friction. The C2M aggregation architecture depends on rapid team closure to trigger manufacturing batches; delays of even hours scatter the demand concentration that enables supplier pricing power.
More devastating would be the restriction of social graph data ingestion. Pinduoduo’s recommendation engine relies on WeChat’s sharing metadata to map influence patterns, nowing not merely that User A shared to User B, but that User B subsequently purchased, and that Users C through G ignored the link. This data flows through API permissions that Tencent grants as a business partner but could revoke as a competitor. Without visibility into the conversion rates of specific social edges, the platform cannot optimize its “team leader” bonuses or target the high-influence nodes that drive viral acquisition. The decentralized, trust-based merchandising collapses into indiscriminate broadcasting, spamming friends without the precision that makes the spam tolerable.
The platform’s countermeasures, investing in standalone app usage, building proprietary social features, face the structural gravity of Chinese mobile behavior. Users maintain five hundred active WeChat chats; they maintain perhaps three active shopping apps. The switching cost of moving team formation from the chat group where the family already coordinates dinner plans to a separate app’s messaging feature is not merely technological but anthropological. The family chat group is the modern Chinese kinship structure; asking it to migrate platforms for the sake of cheaper garlic is asking kinship to reorganize around commerce. It will not happen at scale.
The Auction Floor at the Bottom
The merchant advertising engine faces a subtler stress test as the platform scales toward saturation. The 0.6 percent transaction fee creates a supply base of approximately ten million merchants, many selling identical commodities, white socks, phone chargers, rice, at indistinguishable price points. Visibility within this long tail is allocated through advertising auctions where merchants bid for search ranking and in-feed placement. The economics only function if the marginal merchant can afford to advertise while maintaining profitability; if the auction clears at prices that exceed merchant margins, the supply base exits, and the “treasure hunt” inventory vanishes.
Consider the specific case of a merchant selling disposable kitchen gloves. Manufacturing cost: 1.2 RMB per package. Platform transaction fee: 0.006 RMB. Shipping subsidy: 2 RMB (absorbed by platform). To sell at the 9.9 RMB price point necessary for team-purchase virality, the merchant must acquire customers through advertising at a cost below 1.5 RMB per conversion to maintain a 10 percent net margin. As the category saturates, with ten thousand identical glove sellers, the cost-per-acquisition in the ad auction rises toward 2 RMB, eliminating profitability. The merchant faces a prisoner’s dilemma: exit the auction and lose all visibility, or remain and sell at loss to maintain ranking algorithms, hoping for organic discovery that never arrives.
This dynamic creates a selection pressure toward counterfeit and substandard goods. The merchant who substitutes 0.8 RMB manufacturing cost gloves, thinner plastic, toxic additives, can sustain higher advertising bids while maintaining the 9.9 RMB price point. The platform’s quality control, which relies on user reporting and algorithmic sampling, cannot inspect the ten million SKU long tail with sufficient granularity to prevent the race to the bottom. When the substandard gloves arrive and tear during use, the customer attributes the failure not to the anonymous merchant but to Pinduoduo, degrading the platform trust necessary for social sharing.
The advertising revenue model thus depends on a precise calibration of merchant desperation: sufficient competition to drive bidding wars, but not so desperate that merchants degrade quality to survive. As macroeconomic conditions squeeze small manufacturers, rising input costs, energy price volatility, the margin compression forces more merchants toward the quality cliff. The platform responds by increasing ad load, showing more sponsored listings per screen to maintain revenue growth against declining merchant health. The user experience degrades; the toilet paper seeker finds their search results cluttered with irrelevant sponsored listings, the friction of discovery rising toward the point where the discount no longer justifies the cognitive tax.
The Governance of Variable Cost
Colin Huang’s departure from operational management in 2020, and his subsequent resignation from the board in 2021, functions as an economic variable distinct from corporate governance theater. Huang’s strategic insight, the recognition that social capital could be treated as a variable cost rather than fixed infrastructure, required a specific organizational culture capable of executing the necessary social engineering. The engineering teams had to optimize not for conversion funnels but for “viral coefficients,” measuring not how many users bought but how many users recruited. This metric, alien to traditional e-commerce, required leadership willing to prioritize social graph penetration over immediate monetization.
The transition to Chen Lei’s management introduces a different economic logic: operational efficiency and compliance optimization. Chen’s background in supply chain and risk management shifts investment toward the Duo Duo Grocery logistics infrastructure and away from the experimental gamification mechanics that Huang’s team pioneered. This reallocation reflects a maturation of the business model from growth-at-all-costs to margin defense, but it risks depreciating the very mechanisms, social virality, entertainment retention, that created the user base now being monetized.
The governance structure, with its dual-class shareholding that concentrates voting power with founders despite minority economic ownership, insulates these strategic shifts from activist pressure. This insulation is economically rational only so long as the founding vision remains correct; if the assumptions about social sharing tolerance or gamification efficacy prove outdated, the governance structure prevents the market from forcing strategic pivot. The company becomes a vessel for the specific temporal moment of Chinese social commerce: moment that may pass while the capital structure remains locked in amber.
The variable cost structure for customer acquisition, which enabled the six-year sprint to 788 million users, assumed inexhaustible social graph capacity. Each new cohort of users, however, enters a social environment already saturated with team-purchase requests. The grandmother in the fourth-tier city has already been solicited by her entire extended family; the marginal utility of her next share declines toward zero. The model approaches the asymptote of social exhaustion, where the discounts required to overcome sharing fatigue exceed the manufacturing savings generated by demand aggregation. At that boundary, the engine stalls: the teams fail to form, the C2M batches shrink, the supplier pricing power evaporates, and the business reveals itself as having rented a customer base it never truly owned, holding it only so long as the social fabric remained willing to be mortgaged for the price of cheap fruit.
The exhaustion manifests not as dramatic deletion but as silent recalibration. A user who once maintained active participation in twelve family team-purchase groups gradually mutes the notifications, responding only to direct requests from immediate family members. The peripheral cousins, the former classmates, the neighborhood committee members, these weak ties that once provided the viral transmission vectors, fall dormant. The platform’s algorithm, detecting the reduced sharing velocity, increases the discount incentives to reactivate the behavior, but the user has calculated that the marginal 3 RMB savings no longer justifies the social labor of solicitation. This is the default on the social mortgage: not a refusal to pay, but a recognition that the principal, the relationship itself, has appreciated beyond the interest rate of the discount.
The demographic implications compound this decay. The platform’s early growth exploited the dense social networks of users in their forties and fifties, cohorts with extensive extended family obligations and lower sensitivity to digital privacy norms. As these users age into their sixties, their purchasing patterns shift toward medicinal products and away from the viral agricultural commodities that drove initial engagement. Meanwhile, younger users entering the platform inherit social graphs fragmented by migration and urbanization, ollege classmates scattered across provinces, nuclear families disconnected from agricultural hometowns. The team-purchase mechanism assumes geographic proximity in social relationships; when the friend who might share a fruit purchase lives three time zones away, the twenty-four-hour collection window becomes logistically impossible. The social fabric, once dense enough to support the weight of continuous commercial solicitation, thins into the diffuse, asynchronous connections of contemporary urban life, and the variable cost structure reveals its fatal limitation: there is no cheaper substitute for the exhausted social capital of an aging user base.