China Durian Supply Chain
Seat System White Paper

Industry Background: An RMB 80 Billion Market Without a Responsible Chain

The wholesale era gambled on market swings; the contract era settles supply by covenant. Before discussing institutions, let us first be clear about where this industry stands — durian's first half relied on the category dividend; what will the second half rely on?

1.1　A Decade of Expansion: A Category Grows into an RMB 80 Billion Market

Durian has been one of the fastest-growing categories among China's imported fresh produce over the past decade. According to statistics of the General Administration of Customs of China (GACC), China's fresh durian imports in 2015 were approximately 299,000 tonnes; by 2025, full-year fresh durian imports had reached 1.868 million tonnes, with a value of approximately USD 7.49 billion — over 5,000 tonnes entering the country on an average day. Xinhua News Agency, citing an industry report in June 2026, projected that China's fresh durian imports in 2026 would approach 2 million tonnes, the retail market size could exceed RMB 82 billion, and per-capita annual consumption would be about 1.32 kg.

In ten years, import volume expanded roughly sixfold, and durian moved from a niche, pricey item in tropical-fruit shops to a regular hero category in supermarkets, e-commerce and community stores. This is a solid achievement of China–ASEAN agricultural trade, and the result of joint efforts across logistics, tariffs and cold chain. But the question this white paper discusses begins precisely after scale peaks.

1.2　Inflection Signal: Volume Up, Value Flat — Monthong Goes Commodity

The 2025 data hide a clear signal: full-year import volume grew about 19.7% year on year, while import value grew only about 7.1%. Volume rose by nearly a fifth; value rose by a rounding error — average landed prices are declining. In the first half of 2026, fresh durian imports were about 1.07 million tonnes, up about 47% year on year; the volume surge keeps accelerating, yet prices keep falling.

The driver of falling prices is institutional: under the RCEP framework, tariffs are being progressively reduced, and cold-chain trains such as the China–Laos Railway “Lancang-Mekong Express” have become regularized, sharply driving down the landed cost of Thai Monthong — rail cold-chain transport costs are about 60% lower than road transport, and transport loss has been compressed from as high as about 30% in the sea-freight era to within 3%. The advances in logistics and tariffs have eliminated the location advantage: the old way of profiting by “whoever is closest to the port gets the goods” has systematically disappeared.

1.3　The Location Advantage Is Gone — But the Quality Gap and Trust Gap Are Undiminished

The location advantage can be eliminated by railways and tariffs; the quality gap and the trust gap cannot. The conclusion of 21st Century Business Herald's September 2026 survey of industry players such as Joy Wing Mau is that, over the long term, China's durian market will move toward polarization and structural stratification — the era of uniform market prices and everyone racing to the bottom is over: ordinary durians from Vietnam and Cambodia continue on the low-price route serving mass affordable consumption; old-tree GI production areas in Thailand, tree-ripened and premium Grade-A fruit maintain premiums through stable taste and traceability; Malaysian Musang King and Black Thorn occupy the high end.

This means profit and trust are migrating together toward “quality certainty”: commodity Monthong becomes a bulk good competing on price, while the scarcity of good tree-ripened fruit rises rather than falls. Whoever can stably deliver “this fruit as good as the last one” can stand firm in the stratified market. For terminal merchants this is both an opportunity and a challenge — the challenge being that certainty cannot be manufactured by a single store; it can only come from institutional arrangements on the supply side.

The channel side's response confirms this direction. Category breakdowns of the community group-buying industry show that mass traffic-drivers such as apples, bananas and citrus carry gross margins of only about 15%–25%, while mid-to-high-end seasonal fruits such as strawberries, cherries and durians can reach 40%–50% — the channels' true profit items; the presale model of purchasing against sales can compress loss to within 5%, far below the industry average of 15%–30%. Durian's role is shifting from “traffic gimmick” to “hero category carrying both profit and trust” — but the precondition for a profit item is stable quality, and the precondition for stable quality is contracted supply.

The consumer side is changing as well. Fruit picked early for long-haul transport carries a “blind-box” quality; once the novelty rush recedes, repeat purchase belongs only to channels that are “steady every time the fruit is opened.” Quality certainty is becoming durian retail's only long-term repeat-purchase reason — which is also the fundamental reason this white paper writes “certainty” into its institutional design.

1.4　The Terminal Ledger: A RMB 3-per-jin Margin Cannot Survive One After-Sales Claim

The pressure of market stratification lands first on terminal stores. In late August 2026, Lingshou Media and Huxiu conducted a round of first-hand research on Beijing fruit retail terminals. The ledger of Chen Qiang, owner of a mid-to-high-end fruit store, is a typical sample of durian retailing today (1 jin = 500 g):

• Razor-thin margins: durians are purchased at RMB 20/jin and sold at RMB 23/jin — a margin of only RMB 3 per jin; a five- or six-jin fruit yields a book profit of just RMB 15–18, before deducting labor, rent and utilities.
• Moisture loss means loss: a 30-jin carton purchased at RMB 30/jin (cost RMB 900), fully sold at RMB 32/jin, yields sales of RMB 960 and a book profit of RMB 60, a margin of about 6.25%; after one day on display it loses about 1 jin to dehydration, so even if fully sold the profit drops to about RMB 28 and the margin falls to about 3%. A carton usually sells out in three or four days; the slower the sale, the more visible the loss.
• One after-sales case eats several fruits' profit: when one flesh segment is found half-raw after opening, stores usually cannot compensate just that segment — the whole fruit must be written off; a five- or six-jin durian costs about RMB 100–120 to purchase. “Selling one normally earns a dozen-odd RMB; losing one costs tens or even over a hundred RMB.”

This is not an isolated phenomenon. On listed-company calibers, the industry leaders are likewise contracting: Pagoda Group's FY2025 annual results announcement shows 2025 revenue of RMB 8.174 billion, down about 20.4% year on year, a gross margin of 7.3%, average inventory turnover of 14.3 days (12.6 days in 2024), and 4,468 stores; on the 2025 interim-report caliber, store count shrank from 6,025 in June 2024 to 4,386.

Meanwhile, models focused on the durian hero category are also being validated: Haoshihui, a Guangdong fruit chain with thirty years of history, launched OK Durian themed stores, cutting SKUs from over a hundred to about 40, with durian-related categories accounting for about 40%; a single store stably handles 600–800 orders a day, breaking 1,000 at weekend peaks, with single-store performance about 3 times that of traditional fruit stores. Its founder's summary is that “the core of a fruit store is turnover: sell fast to stay fresh, and only freshness brings repeat purchases,” while warning that influencer-style durian stores have “red-hot front-end traffic but absent back-end control, and are ultimately dragged under by spoilage.”

1.5　Summary: The Category Dividend Peaks; the Institutional Dividend Opens

Putting this chapter's facts together, the conclusion is clear: China's durian market used a decade to complete category penetration — the volume dividend is peaking — prices falling, turnover slowing, leaders contracting, single-store profits knife-thin. At the same time, quality stratification has begun: the scarcity of good tree-ripened fruit and terminals' demand for certainty are rising together. The market does not lack durian; what it lacks is a chain that can take responsibility for the last bite.

This is precisely the question the white paper must answer: where does certainty come from? To answer it, we must first see clearly why the old system cannot provide it.

The Old System's Malady: Why the Five-Tier Distribution Chain Is Fit Only for “Gambling”

Markets are gambled, quality is chanced, after-sales are argued — this is not anyone's malice; the old system is fit only for gambling. This chapter only diagnoses; it does not accuse.

2.1　The Five-Tier Distribution Chain: A Fruit's Journey Through Hands

A durian from Malaysia or Thailand, reaching a Chinese consumer, travels through four to five tiers on the mainstream path: origin orchard → exporter / container buyer → tier-1 wholesale market (large agricultural wholesale markets such as Guangzhou Jiangnan, Jiaxing and Beijing Xinfadi) → tier-2 distribution / regional wholesale → fruit stores, supermarkets, e-commerce → consumer. Every link earns a price difference; no link is responsible for final quality.

Figure 2　Comparison of the old five-tier distribution chain and the Seat System's three-tier short chain. The old chain's markup range is an industry-research interval compiled from the Huxiu / Lingshou August 2026 terminal survey and prevailing industry calibers; the new chain is a schematic of the Seat System's institutional design, and performance capability is subject to Chapters 5 and 6.

2.2　Markups Without Accountability: Risk Like a Game of Pass-the-Parcel

The operating logic of the five-tier chain is “cash on delivery, no exchanges once you leave the door.” Each link earns a resale margin but bears only its own link's inventory risk — once the goods change hands, quality responsibility changes hands with them. One trader has done the math: a 16-tonne batch can cost about RMB 400,000, and “if it sells poorly you cannot recover your capital — this business is one big gamble.” Container buyers gamble on landed markets, wholesalers gamble on turnover speed, stores gamble that stock will not rot in their hands — every rational actor on the entire chain is playing the optimal solution of spot-market gaming.

This structure inevitably produces three behaviors: first, early picking and forced ripening — long-haul spot circulation demands “shelf endurance,” so picking at seven- or eight-tenths ripeness and yellowing on the road is “safer” than natural tree ripening, at the cost of sugar and aroma compounds never having developed; second, passing off inferior goods — the prevailing AA/A/AB/B grading is merely trade convention, not a unified national quality standard; different buyers use different calibers, and the label depends entirely on the seller's word; third, hoarding to gamble on rises — withholding stock when markets rise, stampeding to sell when they fall, amplifying price volatility tier by tier.

2.3　Spoilage and Bad Fruit: A Ledger That Sinks to the Bottom

The second ledger of the spot chain is spoilage. Industry operating calibers show that among supermarket fresh produce, fruit loss rates are about 8%–10% and vegetables about 10%–15%; berry losses can reach 25%–45%, stone fruits and tropical fruits such as cherries, lychees and longans sit in the 20%–35% range, and durian — with its thick shell, sparse flesh, cracking losses and inedible dead packs — is classified as a “high-risk category”; on the consumer side there are even reports of a 6–8 jin fruit yielding only 1.3 jin of flesh.

The frightening thing about spoilage is that it sinks to the bottom of the chain: dehydration, cracking and ripening losses at the wholesale stage ultimately show up in the quality of goods stores receive; while the after-sales losses when stores open dead packs or raw packs are shared by no upstream link. On the consumer side it is a “blind-box experience” — fruit picked early for long-haul transport carries blind-box properties both online and offline, and only at the moment the shell is opened does one know whether the gamble paid off; cumbersome evidence procedures and nowhere to claim are the norm. The flood of consumer feedback on social platforms — “a 6–8 jin fruit yielded only 1.3 jin of flesh,” “the merchant refuses to acknowledge a dead pack” — is essentially the same thing: quality risk at the chain's end has no insurance mechanism.

It is worth noting that spoilage is not impossible to compress — 21st Century Business Herald's survey records a comparison: in the sea-freight era durian transport loss once ran as high as about 30%, while the China–Laos Railway “Lancang-Mekong Express” cold chain compressed transport-stage loss to within 3%. Transport loss can be solved by infrastructure, but the “liability loss” — whether the last bite tastes good, and who pays when something goes wrong — cannot be solved by building railways; it requires a responsible party at the contract level.

2.4　The Terminal's Wager: Squeezed by Inventory, Markets and After-Sales

Chen Qiang's ledger from Chapter 1 is in fact an inevitable result within the five-tier chain structure. Terminal stores simultaneously bear a triple squeeze:

• First, inventory ties up cash. Spot purchasing requires whole-carton, cash settlement; a carton costs nearly RMB 1,000, and if it does not sell out in three or four days, dehydration and ripening eat the entire margin — the slower the sale, the bigger the loss.
• Second, buying gambles on markets. Wholesale prices follow the market, changing daily. Owners keep watch at the market at three in the morning to grab goods — grabbing for price, chancing on quality; “when prices fall, sales do not necessarily rise; when sales rise, profits do not necessarily rise.”
• Third, after-sales with no backstop. One half-raw segment means the whole fruit is written off; upstream wholesale operates “no exchanges once you leave the door,” and the near-hundred-RMB loss is borne by the store alone. Low-price traffic-driving also forms a death spiral: price cuts → over-purchasing → longer inventory → rising spoilage, while the price-sensitive customers drawn by RMB 9.9/jin are not loyal.

Even supermarkets' tiered-pricing playbook (low-price items for traffic, mid-tier for volume, high-quality items for profit) must force on-site fruit opening, leaving after-sales risk at the store; and as for how much linked purchasing and basket-size lift durian traffic actually brings, even supermarket procurement chiefs admit “there is no corresponding data yet.”

The terminal's dilemma in one sentence: however well a single store operates, it can only optimize its own link — it cannot change the upstream spot game. A fruit store can go to the market at three in the morning to pick goods, can drill its fruit-picking craft, can perfect on-site opening service — but which wholesale tier supplies what goods, at what price, and who acknowledges a problem, is not one item within the single store's control. The number of stores operating fruit businesses nationwide is vast (industry-consulting estimates run to the order of several hundred thousand; calibers differ widely and are offered only as an order of magnitude), and the vast majority are independent small stores just like Chen Qiang's. What they need is not one more wholesale market, but an upstream that can carry the three things — markets, quality and after-sales — for them.

2.5　Diagnosis: Not a Moral Problem, an Institutional One

At this point the diagnosis must be stated with restraint and precision: early picking, adulteration, hoarding and blame-shifting are not because people at any one link are bad, but because the old system is fit only for gambling. In a one-shot, zero-sum spot transaction relationship, the responsible party loses instead — whoever does tree-ripened fruit has more “shelf-intolerant” fruit and higher spoilage; whoever pays claims first has higher costs than peers. When no one is responsible for the last bite, gambling is not a choice but a structure; opening a few more stalls or building a few more markets solves the trading-venue problem, not the responsible-party problem.

The old system's disease is not that the chain is too long, but that there is no “responsible party” on the chain — no link is rewarded or punished based on whether the fruit ultimately tastes good. To cure this disease, what is needed is not longer wholesale markets but a new contractual relationship: let the party closest to the trees, most versed in standards and most capable of bearing risk sign the responsibility down on paper.

This is why the Seat System enters the stage. Its complete definition, its separation from models such as franchising, and its institutional precedents in mature industries — including their failure lessons — are the subject of the next chapter.

Defining the Seat System: A Distribution Contract That Vests Capacity Rights

A seat is not a quota spot; it is signable certainty. This chapter gives the complete definition of the Seat System, a written clarification of what it “is not,” and its three institutional precedents in mature industries — including their failure lessons.

3.1　Definition: What the Seat System Is

3.2　What the Seat System Is Not: Legal-Nature Clarification

This white paper may be cited repeatedly in business settings and may serve as a pre-contract document. To prevent any misreading, this section sets out the legal nature of the Seat System item by item in written language.

Comparison with franchising, agency and ordinary membership

3.3　Precedent One: Bordeaux En Primeur — Allocation Works, and It Can Fail

The Seat System is no off-the-top-of-the-head invention. In industries that are “scarce, non-standard and strongly cyclical,” mature markets long ago grew allocation contracts. The most classic sample is France's Bordeaux en primeur system.

How it works: among Bordeaux's roughly 5,000 châteaux, only about 200 top estates participate in the en primeur system. The system runs on the three-tier distribution network known as La Place de Bordeaux: châteaux → courtiers (brokers who advise opening prices and match supply with demand) → about 300 négociants (merchants) → retailers and collectors in roughly 170 countries worldwide; barrel tastings take place in late April, opening prices are released progressively through May–June, and wine is released in tranches, with the first tranche at the best price; buyers pay before the wine is bottled, with delivery 18–24 months later. Allocations are tied to long-term purchasing loyalty: the rarest flagship wines are obtainable only at the en primeur stage; according to wine-trade media, merchants must take up fixed annual allocations, and those who refuse to take goods because prices are too high may have their future cooperation eligibility cancelled — an allocation is not something money alone can buy; it is the present value of a long-term relationship.

But the white paper must write out the other side as well. Of the five vintages from 2018 to 2023, four had en primeur issue prices higher than their later secondary-market prices — the investment logic of “locking in scarcity at a discount” failed; the 2024 vintage met the weakest demand on record — of the 212 châteaux that had previously sold out their allocations in full, only about 50 managed to do so in 2024; the flagship estate Château Latour announced its withdrawal from the en primeur system back in 2012. The repair conditions offered by professional merchants are: opening prices need to be 25%–35% below the expected market price 24 months later, and allocations should be tilted toward genuine cellaring demand rather than speculative resellers.

3.4　Precedent Two: Membership Retail — Costco's Flywheel and Sam's Club's Warning Bell

Costco's flywheel: in FY2025 (ended August 31, 2025), Costco posted net sales of USD 269.9 billion, net income of USD 8.1 billion, and membership fee revenue of USD 5.323 billion; it had about 81 million paid members, with renewal rates of 92.3% in the US/Canada and 89.8% worldwide; it operated 914 warehouses with only about 3,700 active SKUs and an average merchandise margin of about 11%. It deliberately compresses merchandise margins to “barely profitable after costs,” with profit coming mainly from membership fees (about half of profit) — the better the merchandise value, the more stable renewals are; the more stable renewals, the more confidence it has to press supply-chain prices and lock goods. The membership fee is not a ticket; it is the annual rent of a “certainty relationship”; the renewal rate is this machine's dashboard.

Sam's Club China's explosion and warning bell: as of July 31, 2026, Walmart China operated 67 Sam's Clubs with over 10.7 million paid members; at its two annual-fee tiers, membership fee income approaches RMB 3 billion a year; FY2026 omni-channel sales broke RMB 140 billion. The third million members took only 9 months. But the risk side must likewise be recorded truthfully: under high-speed expansion, the supplier audit cycle was at one point compressed from 90 days to 45 days; on June 15, 2026, the State Administration for Market Regulation (SAMR) met with Sam's Club China headquarters over food-safety issues, citing “systemic food-safety hazards across the full chain.”

Together, these two cases illustrate one thing: the certainty created by membership systems must be continuously honored through quality control and after-sales; once quality control fails, the backlash of member trust is fiercer than in ordinary retail. This is why the Seat System ranks guaranteed after-sales as its first foundation (Chapter 5).

3.5　Precedent Three: Contract Farming — Why the Same Institution Mostly Failed

“Locking capacity, locking relationships, preselling allocations” has been invented repeatedly in agriculture — and has failed repeatedly. The white paper does not shy away from these failures.

The failure side has hard evidence: per agricultural science-communication calibers, the national fulfillment rate of contract farming in agriculture is only about 20%. Typical cases include: around 2021, when lemon drinks boomed, some companies organized small farmers to replant at scale under “order purchase” arrangements; after demand slid in 2023, companies cut orders or even vanished, lemon purchase prices fell to RMB 1–2 per kilogram, and bandwagon-following small farmers were forced to cut down their trees; in Luoyang, Henan, over a hundred villagers once signed to plant more than 300 mu of coffee beans, but after harvest the company purchased only a small amount, citing a broken capital chain. In July 2025, the Ministry of Agriculture and Rural Affairs, the Ministry of Public Security and the State Administration for Market Regulation jointly issued an alert against “contract farming” scams, naming such tactics as shell companies fabricating bases, promising buyback prices clearly deviating from the market, and deregistering and vanishing after disputes. An even more extreme mutation is the Ponzi-style fund scheme: a certain group raised funds through a wellness-resort membership and used winery assets as backing; in December 2024 Shanghai police opened a criminal investigation for suspected illegal fundraising, involving over RMB 10 billion and affecting over 200,000 investors.

Taking the failure cases apart, there are three root causes: first, no one bears the price risk — when market prices are high farmers default to sell high, and when prices are low companies reject at downgraded prices citing “substandard quality” or abscond, and contracts cannot hold back the market; second, vague clauses and asymmetric rights and liabilities — quality standards say only “premium” with no metrics, farmers pay high penalties for default while companies default by merely returning deposits; third, high rights-protection costs — scattered farmers cannot afford to sue companies, and companies suing scattered farmers is likewise uneconomical, and in judicial practice the division between natural risk and collaborative-performance liability is often the most wrangled-over link. The Seat System must design in the exact opposite direction on each of these three: price mechanisms agreed in advance, standards upfront and testable, and the company putting heavy assets and after-sales reserves on the table.

The success side likewise has hard evidence: why did Mixue's contract farming succeed? First, heavy-asset binding — it built over 20 standardized planting bases and five production bases in core ingredient regions (total area about 790,000 m²), so the demander's “loss from running away exceeds the loss from performing”; second, guaranteed purchase at the “higher-of” price — where the agreed price exceeds the market price the agreed price applies, and where the market price exceeds the agreed price the market price applies; third, standards upfront with quality premiums — rapid pesticide-residue testing plus third-party re-inspection, AI sorting, with only compliant fruit entering the production line, so good fruit sells at good prices; fourth, long-term contracts and trustworthy settlement — the Guang'an 6,000-mu base signed a 5-year long-term contract; fifth, returns come from a real consumption closed loop — from January to October 2025, over 1.5 billion cups of iced fresh lemonade were sold, with 28 warehouses supporting 12-hour reach to 90% of counties.

3.6　Summary: An Open Institutional Experiment

The Seat System benchmarks the allocation loyalty of en primeur, the relationship contract of membership retail and the scheduling binding of contract farming, while designing in reverse against all three's failure lessons: profit concession written into the price mechanism, allocations given only to genuine operators, standards upfront with backstopped performance, and all returns coming from genuinely selling fruit. It is certainly Dafei's institutional design, but the white paper is willing to hand it over as a public sample: an institutional experiment in China's durian distribution moving from spot speculation to contracted supply — associations, scholars, peers and media are welcome to inspect and question it.

As for “why Dafei should run this experiment” — the two already-solid foundations (guaranteed after-sales and the Super Factory) are unfolded in Chapters 5 and 6 respectively; this chapter first addresses the institution's clauses themselves: the three seat tiers and the Six-fold Certainty.

Three Seat Tiers and Six-fold Certainty: Terms, Boundaries and Fees

Every sentence of this white paper must have a corresponding clause in the contract; nothing is written that is not in the contract. This chapter writes the seat structure, fee nature, Six-fold Certainty and exit mechanism item by item in verifiable clause language.

4.1　The Three-Tier Seat Structure: Boundaries at Every Tier, Performance at Every Tier

Seats are divided into three tiers by geographic scope. The three tiers are a division of service grades, not a division of developmental levels — there is no upline-downline relationship between seats and no headhunting-based compensation.

Figure 4　Schematic of the three-tier seat structure. National total: 3,244 seats (capacity basis in Chapter 7). The above is the public tier description as of the publication date; specific service lists, regional boundaries and quotas are subject to the Formal Contract.

Boundary clauses for regional quota protection

4.2　The Seat Service Fee: Four Service Considerations and the “Three Nots”

One point needs special note: nowhere in this white paper do any supply-price figures or earnings projections appear. The core message on the price mechanism is “prices foreseeable, volatility manageable” — a shift from spot-market pricing-at-market to contracted prices combined with lock-in scheduling; specific prices and settlement methods are subject to the Formal Contract.

4.3　Six-fold Certainty: Benefit Content — Performance Method — Breach Handling

The Six-fold Certainty is no advertising line; each item corresponds to contract-clause language. The table below is a clause overview, with item-by-item boundary notes after it.

Item-by-item boundary notes on the Six-fold Certainty

4.4　Seat Term, Exit and Transfer

Behind the Six-fold Certainty stand two foundations: guaranteed after-sales (Chapter 5) and the Super Factory (Chapter 6). Without the former, a contract has no trust; without the latter, a contract has no capacity.

Foundation One: Guaranteed After-Sales — The Factory End Carries the Risk

Someone has to carry the risk. The factory is closest to the trees — so the factory should carry it.

5.1　2019: The Industry's First Durian Guaranteed After-Sales Policy

In 2019, buying durian was still like opening a blind box; for dead pack, bad fruit and unripeness, the industry's prevailing treatment was “no exchanges once you leave the door.” That year, Dafei took the lead in publicly launching the industry's first durian guaranteed after-sales policy: dead pack, bad fruit, unripeness — fully compensated; flesh yield guaranteed; customers claim compensation with fruit-opening evidence. The policy has run continuously from 2019 to this day and has extended from the retail end to the full chain; it is now written into the contract system as a fixed component of seat services.

The underlying logic of launching the backstop is not charity but quality control: after-sales is the best QC officer. When the cost of bad fruit is borne by the factory rather than sinking to stores and consumers, the factory side has the strongest incentive to perfect every link — harvesting, sorting, cold chain. For the Seat System, this is the first cornerstone of trust — without the history of paying claims from 2019 to today, “the factory carries the risk” is an empty phrase.

5.2　Adjudication Standards: The SCA Defect Dictionary, Visible to All Three Parties

For a backstop to be sustainable, the precondition is that “what is compensated and how it is adjudicated” is public in advance. The Dafei SCA system maintains an official 28-category defect dictionary that clearly distinguishes three situations: fatal bad fruit (inedible situations such as dead pack and severe rot), transport damage (damage caused in the cold-chain and handling links), and physiological blemishes (non-fatal blemishes occurring during growth). Arrival acceptance is adjudicated on site according to the defect dictionary; the dictionary is public and visible to all three parties, serving as the unified basis for after-sales liability division and dispute determination.

The evidence method is likewise put in writing: one fruit one code traceability, traceable to the orchard plot with picking time recorded to the hour; temperature-control logs traceable throughout; photos are taken and retained according to the defect dictionary on arrival / at signing. Partners submit claims with evidence such as fruit-opening photos or videos, and after verification the matter is handled per the agreed method.

The skeleton of the claims process is: arrival / signing acceptance → determine the situation and liability attribution per the defect dictionary → submit evidence (fruit-opening photo or video, traceability code, temperature records) → factory verification → replenishment, exchange or compensation settled at the corresponding goods' value per the rules. The meaning of “guaranteed flesh yield” is likewise defined in the after-sales rules: the flesh-yield index serves as a sorting and acceptance benchmark, and eligible under-yield situations are handled per the after-sales rules — it is a component of the guaranteed service, stated separately from the sorting-standard clause, and does not constitute a per-fruit guarantee for any individual fruit.

5.3　Financial Provision and Single-Batch Pilot Data

The backstop is no verbal promise. The company's cost structure includes an after-sales compensation reserve line item, so the backstop has a budget source; compensable situations, evidence and handling methods are written into the after-sales rules appended to the contract and executed per the rules.

On after-sales performance, the white paper cites only one recorded figure, strictly bounded in caliber: in one fresh-fruit sales pilot, 2,000 Musang King fresh durians sold out within 0.01 seconds, with 28 after-sales cases — an after-sales rate of approximately 1.4%. It must be specially noted that this is single-batch pilot data for fresh fruit, reflecting the actual situation of that pilot batch; it does not represent a perennial average level, still less a promise of future after-sales ratios; the after-sales rate of agricultural products fluctuates with production season, variety, transport distance and the standardization of store operations.

The boundaries are likewise stated truthfully: quantitative SOP details such as compensation response time, crediting time and per-claim caps will be disclosed after signing in the form of the Formal Contract appendix “After-Sales Rules”; before the written standards are finalized, the white paper writes no numerical promises such as “credited within X hours.” The white paper would rather write one point less of promise than let partners make one point more of misjudgment.

5.4　Why the Backstop Holds: A Self-Tightening Closed Loop

The outside world's most common question about the backstop policy is: if bad fruit is fully compensated, will it pay the factory into the ground? Six years of practice give the opposite answer — the backstop is a self-tightening quality-control closed loop: every compensation payment is traced backward through the SCA defect dictionary to the link — was harvesting too early, was sorting missed, or did cold-chain temperature fluctuate? The defect dictionary gives every payment an attribution, and attribution results feed back to revise harvest red lines, sorting standards and cold-chain parameters. The stricter the quality control, the lower the bad-fruit rate; the lower the bad-fruit rate, the more controllable the claims cost; the more transparent the claims data, the more partners and consumers trust.

Three things in this loop are indispensable: first, the reserve mechanism — the cost structure includes an after-sales compensation reserve line item, so compensation is no ad-hoc expense with the panic of no budget ceiling; second, public standards — the defect dictionary and adjudication process are open to partners; compensation is decided neither by the factory nor by the store but by the rules; third, the factory has the capability to reduce problem incidence — the local factory, SCA sorting and the ±1°C cold chain digest quality risk upstream (Chapter 6). All three sit on the factory side; this is why the backstop policy has been continuously executable since 2019, and the confidence with which the white paper dares to write after-sales into seat contracts. The boundaries bear repeating: the continuation and adjustment of the backstop policy are subject to the Formal Contract and the then-effective rules; compensation is limited to the direct loss of the problem goods; and no commitment whatsoever is made on operating income.

Foundation Two: The Super Factory — Capacity and Standards Behind the Seats

Without its own factory and capacity, “lock-in scheduling” is just telling stories with other people's goods.

6.1　The Capacity Foundation: 60,000 Mu of Old-Tree Orchards and a Locally Built Super Factory

Dafei is among the earlier Chinese brands to enter Malaysia and build its own super durian factory locally. The capacity foundation comprises: approximately 60,000 mu (about 4,000 hectares) of Malaysian orchards with trees over 30 years old, commanding a substantial proportion of the region's top-grade old-tree premium capacity; and a locally built Super Factory in Malaysia that holds the full chain — planting, natural ripening, drop harvesting, pre-cooling freshness lock, standardized sorting, quick-freeze packaging and cold-chain logistics — in its own hands. The temperament of tree-ripened durians dictates that real quality control must be moved upstream to the orchard, rather than happening at the wholesale market.

Why must the factory be built at the origin? The reasoning is plain: the golden window after fruit drop is counted in hours, and every hour of delay in pre-cooling, sorting and freshness locking costs a portion of shelf life and flavor; building the factory beside the Malaysian orchards makes “recovered on drop, pre-cooled on recovery” possible, and also means the standard-setter is closest to the trees. The online operating results have likewise passed the market test: Dafei's Musang King products reached the top of JD.com's Musang King category sales ranking (company-statistics caliber); offline, it stably supplies over 4,000 fruit stores nationwide (company-statistics caliber). Without its own factory and channel validation, lock-in scheduling is just telling stories with other people's goods.

6.2　Harvest Red Lines: Ripened on the Tree, Dropped at the Factory

Malaysian Musang King D197 and Black Thorn D200 must ripen naturally on the tree and drop naturally; they may not be picked or force-ripened. The overseas delivery project team is stationed in Malaysia long-term, co-building harvest red lines with the orchards: if the fruit stem has not naturally loosened and the ripening node has not arrived, harvesting is prohibited; fruit is recovered the moment it drops and enters the pre-cooling stage within 30 minutes of harvesting per factory standards, going directly to Super Factory sorting. The golden window after drop is counted in hours, and the multi-hand turnover speed of the spot market cannot keep up with the speed at which fruit ages — this is precisely the meaning of placing the factory upstream.

In the cold-chain link, full-chain temperature fluctuation is controlled within ±1°C, with smart temperature-controlled zoned warehousing; fresh fruit maintains a shelf life of about 10 days after arriving in China, above the industry's common 3–5 day level. The traceability caliber is unified as one fruit one code + temperature-control logs traceable, traceable to the orchard plot and the picking time point.

6.3　SCA: A Unified Language of Evaluation

The old distribution's standards were chaotic: everyone labels “Grade A,” everyone with a different caliber. What SCA (Dafei Durian Grading & Evaluation System) does is turn “by experience” into “by standard”:

• Six-level ripeness determination: tree-ripened, near-ripe, mid-ripe, unripe, hard and early-picked; early-picked fruit shall not enter the tree-ripened supply system;
• 28-category official defect dictionary: the three categories of fatal bad fruit, transport damage and physiological blemishes, serving as the unified basis for sorting and after-sales adjudication;
• Physicochemical index system: indices such as sugar content (°Brix), sugar-acid ratio and dry matter are incorporated into sorting; core reference ranges — Musang King flesh yield of 25%–40%, Black Thorn flesh yield of 30%–45%, tree-ripened sugar content of 38–42 °Brix;
• Five commercial grades: abolishing the market's chaotic custom “A-grade” labels; fruit is uniformly sorted into the factory's five grades, standardized quick-frozen for freshness locking and modified-atmosphere packaged before leaving the factory.

All the above index ranges are the factory's sorting execution standards and the dispute-determination benchmark for arrival acceptance, not a per-fruit guarantee for any individual fruit (boundary clauses in Chapter 4).

6.4　MW and ITC: Delivering the Knowledge System and the Data Foundation

MW (Durian Delivery Knowledge System) crystallizes ten years of experience into trainable, replicable content: a course system covering modules such as regional terroir, variety genetics, harvest windows, freshness-lock processes, spoilage control, after-sales adjudication and tasting-and-sales, supported by online courses, offline certification camps and pre-season scheduling camps, with store staff trained and certified before taking posts; at the same time, origin footage and quality material packs are packaged for WeChat Moments, livestream and short-video scenarios, ready to use out of the box. Fixed training frequency and material-update cadence are subject to the contract service description; continuously iterated content is not a certain-delivery obligation.

The enablement system's design principle is “turning the factory's capabilities into the store's daily routine”: store staff need not become durian experts — the SCA defect dictionary tells them how to judge fruit, MW courses tell them how to present fruit, and the material kits help them sell fruit, while the delivery project team is on call throughout the season. The only thing stores truly need to do well themselves is serve their local customers. This is consistent with the S2b2C enablement logic, and is the fundamental difference between the Seat System and the wholesale relationship of “only supplying goods, not caring about sales.”

ITC (Origin Trade Data System) is the data foundation for lock-in scheduling: the origin daily fruit-drop collection team is already running, the annual production-sales rhythm calendar has landed and is updated yearly, and quarterly season data briefings are continuously delivered to the seats' private-domain communities; on/off-year capacity models and sea/land loss benchmarks are under construction [Planned] and incorporated into quarterly briefing references. It must be stated truthfully that capabilities such as store-dedicated real-time data dashboards and AI shortage / price-hike / poor-fruit alerts remain under continuous construction and gradual rollout and are not written into the existing-capability list at this stage. The usage boundaries of ITC data (no guarantee of absolute accuracy; decisions at one's own risk) are set out in the Chapter 4 data-tools clause.

The three systems have a clear division of labor, and together they constitute the full meaning of a “Super Factory”: SCA solves “is it good, and by what standard is it judged” — the factory's quality-inspection language; MW solves “do they understand it, and can they be taught” — the factory's delivery language; ITC solves “how much is produced, and when to lock” — the factory's scheduling language. In the old distribution these three things were scattered across the picking master's experience, the wholesaler's verbal promises and market conditions no one could pin down; the Super Factory crystallizes them into standards, courses and data, and what seat partners buy is essentially usage rights to these three languages.

6.5　Fulfillment Network: 7 Forward Warehouses and the Timeliness Position

Seven forward warehouses are deployed nationwide to support flexible fulfillment for seats: single-item dropshipping from 1 item; chartered air freight from a 10-carton minimum, full cold chain at ±1°C. The timeliness position is: delivery to warehouse within 24 hours in 108 cities; next-day delivery in 208 cities; delivery to store within 24 hours for 10km-Radius Seats. Fresh-fruit off-season windows (annual windows such as March–April and October–December) are bridged by regular frozen flesh for volume; the extreme off-season is filled by liquid-nitrogen Musang King and liquid-nitrogen Black Thorn — the liquid-nitrogen line uses -196°C liquid-nitrogen instant freshness-lock technology, freezing the flesh cells within the golden window after fruit drop, with year-round constant quality and no seasonal fluctuation, serving premium gifting scenarios. It should be noted that the liquid-nitrogen frozen line bridges off-season supply and premium scenarios and cannot replace the taste of fresh fruit; “year-round uninterrupted fresh-fruit supply” is not and will never be a promise of this white paper.

The significance of the 7-warehouse network is not only speed but also dissolving after-sales risk on the factory side: forward warehouses perform inbound and outbound inspection per SCA standards, temperature logs are fully traceable, and every carton received by stores and consumers carries complete temperature and sorting records available for checking. It must be stated truthfully that the occasionally circulated claim of “36-hour home delivery” is not the company's current verifiable external position — this white paper uniformly adopts the positions of 24-hour delivery to warehouse/store and next-day delivery; any promise of a shorter timeliness shall be subject to the Formal Contract after actual measurement and verification.

Figure 6　The two-foundations support chart. Foundation capabilities are subject to the company's actual operations as of the publication date; capabilities under construction (real-time dashboards, AI alerts, geofencing, quantitative after-sales SOPs) are all expressly marked in the text as “under continuous construction and gradual rollout” and are not listed as existing capabilities.

Quota and Capacity Mechanism: Why 3,244 Seats

Ripen one, lock one, reserve one for you. 3,244 is not a marketing number; it is the physical ceiling set by capacity rigidity.

7.1　3,244 Seats: The Capacity Ceiling That Can Be Committed for Lock-In

The most frequently asked question is: why are seats limited? The answer is plain — the number of seats equals the ceiling of capacity the company can commit to lock-in scheduling. Tree-ripened durians must drop naturally and cannot be force-picked; there are only two major production seasons a year; trees over 30 years old cannot be expanded in the short term; and top-grade old-tree capacity is a physical boundary. The intersection of orchard capacity, factory sorting capacity and forward-warehouse fulfillment capacity yields the committable supply scale, which converted into seats is 3,244. Releasing seats beyond this number would be over-promising to partners.

That is why the white paper uses no phrases like “last spots” or “miss it forever.” The limit is not hunger marketing but the performance sincerity of “not daring to oversell”: the more seats released, the thinner the certainty each seat receives; welding the seat count to the capacity ceiling is precisely the precondition for the backstop to hold and quotas to be honored. The durian distribution market is highly fragmented; the company's share is limited and it holds no dominant market position; the number 3,244 constrains only ourselves and constitutes no market barrier whatsoever.

7.2　Lock-In Rhythm: Scheduling Follows the Drop Data

Tree-ripened fruit “ripens one and drops one”; it cannot be mass-produced or stockpiled, so seat supply operates by locking orders, locking prices and locking quotas in advance. Lock-in scheduling is not a rigid guarantee by the calendar but follows origin data: the collection team continuously gathers daily drop volumes and season intelligence in Malaysia; the ITC system outputs season curves and the annual production-sales rhythm calendar accordingly; along the full-year season curve, quota locking for the next peak season starts in the preceding season window. Malaysia's production seasons may shift as a whole by about ±1 month with drought or flood, and scheduling is dynamically corrected by real-time origin intelligence.

Figure 7　Schematic of the capacity lock-in rhythm and annual timeline. Season start/end may shift forward or backward by about ±1 month with climate; specific scheduling is subject to ITC season briefings and the Formal Contract.

7.3　What About Volatility: Proportional Fulfillment and Force Majeure

The white paper does not evade the possibility of reduced output. Durian is produced seasonally and sold year-round; on/off years, extreme weather, pests and disease, and trade policy can all cause actual available capacity to fluctuate. The mechanism design is:

• Proportional fulfillment — in the event of capacity volatility, fulfillment is arranged proportionally across each seat's agreed quota, with priority to the locked portions; there is no “highest bidder wins”;
• Off-season bridging — fresh-fruit windows and the off-season are bridged by the regular frozen flesh and liquid-nitrogen frozen systems, with no gap in premium scenarios;
• Rules upfront — capacity-adjustment and force majeure situations and handling methods are pre-written in the Formal Contract and fully disclosed before signing; resulting shortfalls are handled per the contract clauses, and the company bears no compensation liability beyond that agreed in the contract.

The time discipline of locking is likewise public to partners: per the current production-sales rhythm, from each September lock-in for the next January's top-grade fresh-fruit peak must start (January 2027 is the next super prime season), while locking for the July–August summer-autumn peak is completed before the season; stray-flowering crops are small in volume and volatile in quality (for example, September 2026 saw a window of scarce fruit, high prices and unstable quality), and scheduling flags this truthfully in briefings rather than passing stray-flowering off as prime-season quota. Ripen one, lock one, reserve one for you — what locking locks is future fruit, relying on present data and discipline.

It must be stated once again: this white paper makes no absolute promise of the “uninterrupted supply under any circumstances” type — matters affected by natural conditions can only be constrained by mechanisms, not backstopped by slogans. This is not conservatism but honesty about agricultural law, and it is the watershed between a seat contract and investment-pitching rhetoric: adjustment rules written in the contract are more reliable than verbal absolute promises.

7.4　Quota Release and Transfer

The 3,244 seats are released directionally by city tier rather than cast across the whole web, with seats mapping one-to-one to regional quotas. Seats may not be transferred, subleased, sold for a price or speculated on by partners (exit and transfer clauses in Section 4.4); where a partner continuously fails to meet the minimum contracted performance standards, the company has the right to adjust the protection arrangements or even reclaim the seat per contract — quotas are reserved only for co-builders who genuinely operate and genuinely sell fruit. This clause is directly benchmarked against the Bordeaux en primeur lesson: once allocations flow to speculators rather than genuine operators, the institution is devoured by its own design.

7.5　Chapter Boundary Statement

Per the white paper's discipline — only what can be honored is promised — all capacity-related statements in this chapter are mechanism descriptions rather than outcome promises: lock-in scheduling is an operating mechanism already running, but “how much capacity is locked and how much production is scheduled” is content disclosed per contract after signing; 3,244 is the currently calculated committable ceiling and will be recalculated on the same basis as capacity changes; fresh-fruit supply fluctuates naturally with the seasons, windows and the off-season are bridged by the frozen system, and shortfalls under extreme conditions are handled under the contract's force majeure and capacity-adjustment clauses. The white paper writes no “grab a seat” or “appreciation” — only capacity, mechanisms and boundaries. The meaning of the limit has never been to manufacture a sense of scarcity, but to ensure that the certainty of every seat already signed is never diluted.

Seat Partner Profiles, Onboarding Process and Co-Building Outlook

Seat partners are not customers; they are industry co-builders. The white paper's final chapter makes clear who suits taking a seat, how to enter, how to exit, and to whom this institution is ultimately to be returned.

8.1　Who Suits Taking a Seat: Partner Profiles for the Three Tiers

There are three one-strike veto items written into the entry standards: first, those who pursue only low prices and do not identify with the tree-ripened philosophy are not accepted; second, those accustomed to competing on commodity goods and rhetoric and unwilling to implement SCA quality-control standards are not accepted; third, those who cannot accept the agricultural premise that “supply fluctuates naturally with the production season” are not accepted. The Seat System serves those willing to treat quality as a long-term business; it does not serve those gambling on markets.

8.2　How to Enter: Entry and Onboarding Process

The onboarding process runs per existing mechanisms, with no fabricated time limits for any stage:

1. Targeted invitation and application — seats are released directionally by city tier; partners submit their operating situation and cooperation intent;
2. Field visits and system experience — field visits to the origin or model stores to understand the orchards, factory, sorting and after-sales systems (experiential arrangements are subject to actual notice);
3. Signing — signing the Formal Contract, specifying seat tier, regional boundary, service list, fees, exit and dispute-resolution clauses;
4. Delivery docking — the delivery project team (five roles: quality control, supply chain, operations, content, training) sets up a dedicated group and opens traceability queries and season briefings;
5. Training and certification — onboarding training and staff certification, unifying SCA adjudication calibers and after-sales processes;
6. First lock-in and first order — confirming the first locked quota, with the nearest forward warehouse arranging the first-order shipment.

The partners' corresponding obligations are equally clear: execute arrival acceptance and after-sales evidence retention per the SCA defect dictionary; return terminal sell-through data as agreed (the data pool guides scheduling in reverse); no cross-territory diversion selling; no selling early-picked fruit or commodity goods that damage the system's reputation. The handling path for triggering red lines is: warning and rectification → downgrade or supply suspension → seat reclamation per contract; specific metrics and settlement rules are subject to the Formal Contract.

Daily docking mechanism: each seat is assigned a delivery project team docking directionally across five roles — quality control, supply chain, operations, content and training — with pre-season camps, in-season daily reports and post-season reviews; quarterly season data briefings and the annual production-sales rhythm calendar are delivered regularly through the seats' private-domain communities, and ready-to-use material packs of origin footage, quality-control stories and quality macro-shots are produced continuously through the seasons. Certain-delivery content (training sessions, material versions, briefing frequency) is subject to the contract appendix “Seat Service List”; iterative content beyond the list continues to advance per the product roadmap — what partners interface with every day is not a salesperson but an interface to a set of factory capabilities.

8.3　How to Exit: Enterable, Verifiable, Exitable

The exit mechanism was set out in Section 4.4; three principles are reiterated here: first, seats run on an annual cycle, and partners may exit under the conditions and procedures agreed in the contract; second, upon exit, fee settlement, in-transit order handling and data-access termination are executed strictly per the Formal Contract, and the white paper makes no unconditional or anytime-refund promise; third, seats may not be transferred or speculated on, and changes of operating entity require a written application and, upon review and approval, handling per contract. A system that only lets people in is not worth trusting; only a system that writes its exit rules clearly can be signed for the long term.

8.4　Co-Building Outlook: From Enterprise Sample to Industry Institution

The white paper is positioned as a distribution-institution proposal, not as investment solicitation or recruitment material. Its endpoint is not signed-contract numbers but the commonization of the institution. The directions we advocate and are willing to open for participation include:

• Co-building group standards — promoting discussion and formulation of group standards such as “tree-ripened durian grading” and “durian guaranteed after-sales service norms,” turning the SCA grading framework, defect dictionary and after-sales rules from enterprise standards into language the whole industry can share;
• Open inspection — opening factory visits, season briefings, sorting records and pilot data samples; associations, scholars, peers and media are welcome to inspect, question and falsify;
• Data for common benefit — industry insights formed from returned terminal sell-through data are shared with the industry while protecting partners' commercial information, reducing chain-wide blind scheduling and hoarding gambles.

China's durian industry's first half relied on the category dividend; its second half must rely on the institutional dividend. The Seat System is the first piece of homework Dafei hands in, but it does not belong to Dafei — if this contract logic of “scheduling + standards + backstopping” can be discussed, revised and adopted by more people, letting durian sellers sleep at night and durian buyers cut without trembling hands, then this white paper will have fulfilled its mission.

Good fruit ripens on the tree; seats are locked at the factory. The wholesale era gambled on market swings; the contract era settles supply by covenant. An institution's worth lies not in the applause at its release, but in whether it still stands when it is dug up ten years later.

Further reading: the white paper's companion essay, The “Seat Era” of Durian Distribution: From Spot Gambling to Contracted Supply (in Chinese), addresses “why change is unavoidable” (click to read); for industry-status analysis, see also Prices Collapsed and People Still Won't Buy: Five Chronic Ills of the Durian Industry and the Path to Breakthrough (in Chinese); more industry research is collected at the Durian Knowledge Blog (in Chinese). The white paper covers “what it is and how it performs”; the essay covers “why” — read together, the full picture of the institution emerges.

Index of Main Data Sources

Industry data are sourced from government bodies, industry associations and public reports, with sources indicated; company operating data are sourced from the company's own statistics. Data are subject to differences in caliber and timing and reflect the situation as of the publication date.

• Import volume and market size: General Administration of Customs of China (GACC) fresh durian import statistics; Xinhua News Agency, “Seeing China's Big-Market Opportunities Through a Single Durian” (June 4, 2026).
• Market stratification and cold-chain loss: 21st Century Business Herald, “Durian: A ‘Sweet Sample’ of China–ASEAN Win-Win Cooperation” (September 4, 2026).
• Terminal operating ledgers: Huxiu, “‘Durian Prices Plunge, Yet Still Won’t Sell’” (August 28, 2026); Lingshou Media (Linkshop) report of the same title (August 31, 2026).
• Leading-chain operating data: Pagoda Group FY2025 annual results announcement (HKEX / CNINFO, March 26, 2026); Shenzhen News Network (August 28, 2025).
• Durian hero-category store model: Jiemian News, “Haoshihui Fruit × Lemeng: One Durian Levers a New Growth Curve” (August 17, 2026).
• Fresh-produce loss reference: SMZDM fresh-produce compensation explainer (2026), citing industry loss reference standards.
• Costco: Costco FY2025 CEO letter to shareholders and financial statements (public reports, September 2025 – January 2026).
• Sam's Club China: 21st Century Business Herald (September 1, 2026); China Business Journal, “Sam's Club China's ‘Troubled Season’” (June 29, 2026).
• Bordeaux en primeur: En Primeur Club, “Bordeaux 2025 En Primeur” (June 2026); The Luxury Playbook (April 2026); Wine Review (June 2026).
• Contract farming: Ministry of Agriculture and Rural Affairs official alert on “contract farming” scams (July 3, 2025); CNR.cn / Dahe Daily (September 2014); NetEase / Neicanjun (January 2025); Securities Times, WBO Wine Business Observer (December 2024).
• Mixue contract farming: Shanghai Securities News (November 19, 2025); China.com.cn (June 15, 2026); Mixue official website.
• Company's own data: the 60,000-mu old-tree orchards, 4,000+ supplied stores, 7 forward warehouses, 108-city / 208-city timeliness, single-batch pilot after-sales rate of 1.4%, etc., are all Dafei Super Durian Factory company-statistics calibers, labeled in the body respectively as “company statistics,” “single-batch pilot data,” and other qualifiers.

Note: market-size forecasts, turnover-day and loss-rate data from third-party consulting firms' templated reports (Docin / Boyan-series reports, etc.) were excluded from the body of this white paper due to mutually contradictory calibers; matters for which no authoritative unified data exist (such as domestic durian output and the standalone durian-category loss rate at fruit stores) are not asserted in the body.