Bản thảo để góp ý, phiên bản 0.6. Không dùng để trích dẫn.Báo cáo này là bản dự thảo và có thể thay đổi.
AltProtein VietnamNghiên cứu phạm vi phía cung tại Việt Nam
EN
  1. Trang chủ
  2. Báo cáo
  3. V. What to do

Chương 25 · Phần V

25. What demand asks of the frontier: product targets, plays and public goods

Demand in Vietnam pulls new protein as an ingredient inside foods people already eat, sold to the firms and kitchens that make them, and soy, not meat, is the price an extender must beat. We turn this into eleven target product profiles, a demand check of the ten plays and twenty-one cheap moves, each with a funder or budget.

24 phút đọcÝ nghĩa các nhãn

Nội dung trang này bằng tiếng Anh. Đọc bản tóm tắt tiếng Việt

Trong trang này
  1. 25.1 Demand findings that change what to make
  2. 25.2 Target product profiles
  3. 25.3 The ten plays, checked against demand
  4. 25.4 Twenty-one cheap moves, and who would pay
  5. 25.5 What not to do
  6. 25.6 What would change our view

In one paragraph. Part III's answer to "what demand exists for new protein in Vietnam?" is clear, even where the evidence is thin. Demand pulls new protein as an ingredient inside foods Vietnamese already eat, sold to the firms and kitchens that make them: chay makers replacing imported textured soy, meat processors extending pork, plant-milk makers adding protein, canteens feeding workers on about VND 25,000 to 31,000 a meal, and exporters of dumplings and other wrapped foods. It does not pull branded replacements for meat. We tested this against 80 named actors and 16 further lines of research (Actor check). The core answer holds, but the route to each buyer is narrower and more concrete than our first reading suggested. The rules on school menus, protein claims, blend labels, tariffs and sugary drinks are already written; soy, not meat, is the price an extender must beat; and domestic textured soy is an import-substitution plant best owned by a chay maker, co-packer or distributor and financed with debt, not a venture bet. This chapter turns that into eleven target product profiles, a demand check of the ten plays of Plays and twenty-one cheap moves, each with the funder or budget that would pay for it.

25.1 Demand findings that change what to make

Ten findings from Part III.

  1. The price to beat is eggs and chicken, at VND 34,000 to 45,000 per 100 g of protein; lean pork (about VND 71,000) is the ceiling for a mass meat analogue. Imported analogues sit at 12 to 18 times eggs and chicken (Protein diet) . VND 34,000 to 45,000 per 100 g of protein
  2. Adults do not need more protein; dishes need denser protein. Protein sits in toppings, fillings and rolls; chay foods have a median of 4.7 g per 100 g against 19 g for lean pork (Protein diet). For school-age children protein is not the gap either: urban school lunches carry about 31.8 g, and upland boarding lunches and dinners carried meat, fish or egg in 240 of 242 meals; the documented shortfalls are breakfast, calcium and vitamin A 1,2,3 . Stunting (18% in 2023, 32% among ethnic minority children) is set before age two, where plant products face the young-child protein-quality pattern and registration as foods for children under 36 months 4,5 .
  3. Few want to eat less meat (15% of urban online respondents, Consumers), and meat per person is projected to rise 25% by 2035 (Protein diet) .
  4. What people ask for is variety, cooking fit, fewer additives and safety, not a closer copy of meat (Consumers) .
  5. The modern plant category that sells is plant milk, priced at or below dairy by local incumbents. The best plant-milk SKUs sell 60 to 79% of the litres of the best dairy SKUs on Lazada and Tiki, and Vinamilk's high-protein nut milk sells as a premium niche of about 70,300 L at its Lazada store 6 .
  6. Chay is the ready channel and the wrong impact count: a shelf, a cold chain, a calendar and makers already buying imported protein, but chay days are already in the meat baseline (Chay baseline). Fake chay is a latent rumour, not a live scandal: we found no confirmed case of meat in a Vietnamese chay product from 2015 to 2026; chay's live trust issues are borax, microbes and unlabelled goods 7 .
  7. Blends are the entry route, and soy already holds it. One large processor lists soy protein in 31% of its meat formulations, and seven commercial Vietnamese sausages contained soy and were as well liked as alternatives 8,9 .
  8. Words carry risk: nhân tạo (artificial) and giả (fake) are loaded, and độn (padding, filler) means hidden cost-cutting; lên men (fermented) helps. A declared soy ingredient in branded sausages draws no observed penalty 10,11 .
  9. Kitchens and factories come before shelves, and they already buy cheap plant protein. Only an ingredient fits a canteen's protein budget of VND 7,500 to 14,000 a meal, and domestic dry soy-gluten pieces already reach kitchens at about VND 4,000 to 5,100 per 20 g of protein, against VND 7,200 to 9,500 for tofu and VND 8,800 for eggs (our calculation from 12) . VND 7,500 to 14,000
  10. The volumes that exist are business-to-business. On the benchmark path the routes create demand for about 19,000 t of protein a year by 2035, of which about 11,000 to 17,000 t made in Vietnam for the home market, mostly as import substitution, and displace about 0.2% of meat protein (Demand sizing) . about 19,000 t delivered; about 2,400 t (0.2%) displaced

Six findings from the actor check that change the picture.

  • Schools buy tofu beside meat, and the rule that mattered has expired. Decision 1340's animal-protein target ended in 2020. The current guidance, Decision 3958/QĐ-BYT (December 2025), sets no animal share and asks for plant protein (legumes, tofu) at least twice a week. Audited menus meet that in only 9 of 48 school-weeks, plant protein appears on 15.1% of 239 lunch days, almost always as tofu beside a meat dish, and the menu software that caterers use prints a standard of "animal protein at least 48%". A plant ingredient in the plant slot competes with tofu and replaces tofu, not meat; the only school route that removes meat is a labelled blend in the minced-meat and fish-cake dishes served on 28.5% of days 13,14,1 . The rule window is Decision 973/QĐ-TTg: a MOET bán trú (school lunch and midday care) circular due in 2026 and digital menu tools due in 2027 to 2028 15 .
  • A sugary-drink excise will shape plant milk. From 1 January 2027 drinks above 5 g of total sugar per 100 ml pay 8% (10% from 2028); dairy is exempt and whether sweetened soy, nut and cereal milks are in scope is unresolved. Core Fami soy milks carry 6.5 to 9.3 g, and the reference high-protein nut milk is also above 5 g 16,17,18 .
  • There is no tariff shelter. Textured protein, soy isolate, pea protein and gluten enter from China, ASEAN and Korea at 0% under ACFTA and RCEP; the 5%, 3% and 10% duties apply only to other origins, and Chinese, French and Canadian pea protein also enter duty-free 19 .
  • Retailer private label is a working route for chay. All five Co.op Select frozen chay dishes are made by one certified co-packer, GN Foods, and Lotte Mart sells its own chay label; the products are starch and vegetable led and carry no protein claim 20,21 .
  • Most label rules are already in force. A high-protein claim needs 10 g per 100 g (TCVN 14429:2025, voluntary), and 20 of 43 audited foods already meet the source-of-protein condition without claiming it; TCVN 13625:2023 already defines vegetarian and vegan claims; any ingredient named in a product name must be quantified; no standard sets a minimum meat content; and the GMO label applies only above 5% of ingredients 22,23,24,25 .
  • Soy is the extender benchmark, not meat. Replacing 20% of the lean in chả lụa (steamed pork roll) with hydrated soy saves VND 9,100 to 16,600 per kg at every hog price since 2019, so a new extender must beat hydrated soy at about USD 2.5 to 2.7 per kg of protein, and win on sensory tolerance or the label, not on price against pork 9,26,27 .

25.2 Target product profiles

A target product profile states what a product must be for the demand that exists. The full profiles, with sources, are in target_product_profiles.csv . TPP-11 is a feed profile for play T2. A zinc-fortified soy dish base for upland school kitchens was proposed and withdrawn: the kitchens it named already serve meat almost every day, so it stays only as a research option for fortified breakfast items 28,5 .

ProfileFirst buyersPrice targetProtein targetMust haveFirst test
TPP-01 Domestic textured soy or soy-gluten proteinChay makers and dry chay-piece makers; co-packers incl. GN Foods; distributors (Brenntag, VMC, TDGIMEX)At or below Chinese landed textured soy (USD 1.29 to 1.38 per kg); delivered soy flour below about USD 700 to 940 per t50 to 70% as soldMade from soy flour or meal; several shapes; non-GMO where asked; debt-financed ownerFive delivered soy-flour quotes; conditional orders of 2,000 t a year from three buyers
TPP-02 Higher-protein chay foodsIncumbent chay brands; Co.op Select (GN Foods) and Lotte Choice L private labelToday's chay shelf (VND 56,000 to 173,000 per kg)10 to 12 g per 100 g"10 g đạm/100 g" on the front; short label; stock for lunar peaksPitch a 10 g version of a Co.op Select SKU; minimart repeat test
TPP-03 Hybrid extender for processed meatVissan, Masan MEATLife, C.P., CJ, SG FoodAt or below hydrated soy (about USD 2.5 to 2.7 per kg of protein)Finished product 10 g or more per 100 g10 to 20% in fine emulsions, up to 30% in coarse mince; plant part named as an ingredientBlind test at equal saving against the processor's soy formula, in xúc xích (sausage)
TPP-04 Canteen protein ingredientCaterers via chay-piece makers and HoReCa distributors; later school, hospital and upland kitchensBeat dry soy-gluten pieces (VND 4,000 to 5,100 per 20 g of protein)About 20 g per adult dish; 10 to 15 g per childBulk-cook tolerant; caterer documents; no children's claimsCaterer trial with extender and school arms
TPP-05 Rice or mung bean isolate for plant milksVinamilk, Vinasoy, Nutifood, THBetween Chinese and Western pea protein (USD 2 to 8.5 per kg), unprotected by tariff80 to 88% isolate; drinks at 5 g or more per 100 mlUHT-stable, neutral, soy-free; host drink at or below 5 g sugar per 100 mlSamples to two dairy research teams
TPP-06 Fungal biomass functional ingredientBakery, seasoning makers, processorsAgainst egg white, dairy proteins and yeast protein, not soyAbout 45% dry basisClean label; binding and umamiTrials benchmarked against yeast protein and egg white
TPP-07 Precision-fermented specialty proteinFormula and medical nutrition makers; exportAgainst conventional lactoferrin and whey concentrateInfant and medical gradeGMO question settled for high-inclusion usesLegal opinion on the one open GMO question
TPP-08 Export plant-based wrapped foodsEU and UK importers and brand owners of Asian wrapped foods; CJ's Kizuna plantNear European private-label pricesProtein claims where trueIFS or BRCGS; EU wrappers within the 20% flour capQuotes from three importers and two brand owners
TPP-09 Diaspora protein-upgraded chayAsian grocers abroadUSD 3.50 to 10 a pack10 g or more per 100 gFrozen logistics; tariff risk pricedListing test with one importer
TPP-10 Cultivated meat or seafoodNone domestic before the mid-2030sNot applicableNot applicableHybrid format; not sold as chayResearch partnerships only
TPP-11 Functional microbial feed ingredient (T2)15 ASC-certified feed mills; importers of feed yeastAt or below imported bulk yeast (about USD 1.1 per kg) unless trials show a gainPriced on performanceHead-to-head data; listed raw material; ASC data1 to 2% tank trial against the mill's current yeast

25.3 The ten plays, checked against demand

The market-pull scores of Plays were set from supply-side evidence of buyers. Part III's evidence suggests two changes and no reductions, and the actor check keeps both while replacing the evidence behind T1 and T2 :

PlayMarket pull, supply viewDemand viewWhy
T1 Domestic textured plant protein45Several named buyers already buy the imported equivalent: textured-type imports from China, India and Serbia were 12,832 t in 2025; export origin rules add no customer 29,30,31
T2 Functional microbial feed ingredients44Vietnam already imports about 9,200 t of inactive yeast a year for food and feed, 95% of it at about USD 1.1 per kg; export standards neither require nor reward microbial ingredients 32,33
T3 Bulk microbial protein for fishmeal33Same pull as T2; price-led 34
T4 Cassava sugar hub22Export demand is for starch; no new protein demand 35
T5 Shared pilot and tolling22An enabler; adds a need for application and sensory trials
T6 Duckweed feed22Not examined by the demand streams
T7 High-value precision-fermentation ingredients22, conditionalBuyers of the target ingredient exist and sell non-GMO; the GMO label bites only above 5% of ingredients 36,25
T8 Cultivated seafood research11Revealed Asian sales are tiny; the regional cultivated-seafood firms have wound up or moved to pet food and skincare 37,38,39
T9 Koji and fungal foods22 (3 if a processor pilots a fungal extender above soy parity)Fermented reads well; the route is blends and chay ingredients, but soy already fills the extender role 40,9
T10 Mung bean and rice protein for plant drinks34Vinamilk buys pea protein for a soy-free high-protein milk that sells as a niche; the sugary-drink excise and duty-free pea protein narrow the room 41,6,17

T1: the market pull stands, the thesis changes. T1 keeps a market pull of 5 because named chay makers already buy the imported product. But the pool a soy extrusion line can serve is about 12,800 t of textured-type imports, not the whole 40,000 t of imported plant-protein ingredients: the near tripling of plant-protein imports is mostly soy isolate and other HS 3504 protein at about USD 1 per kg, which needs other plants 29,30,42 . Capital is not the barrier. Shandong makers list complete dry lines of 0.5 to 1.5 t an hour at USD 11,000 to 100,000, a whole plant installs for about USD 0.3 to 1.4 million, and working capital of USD 0.6 to 3.5 million is the larger need 43,44,45 . USD 0.3 to 1.4 million There is no tariff shelter: Chinese product lands at USD 1.29 to 1.38 per kg, and India-origin textured soy lands lower still. The delivered price of food-grade defatted soy flour, 63 to 85% of cost and imported from the same countries that export the finished product, sets the sign of the margin (our calculation) 30,46,19 . about USD 700 to 940 per t The benchmark path supports about one 1 t/h line, or two 0.5 t/h lines, by 2035. T1 is therefore an import-substitution plant best owned by a large chay maker, a group of chay makers, a co-packer or an ingredient distributor and financed with debt and trade finance, not venture equity: its return is a commodity spread that Chinese exporters can compress, it holds no intellectual property and its exit is a trade sale .

T2: the same score on revealed evidence. Our first reading took T2's pull from export feed standards. Those standards are neutral: ASC v1.2, BAP and GLOBALG.A.P. neither require nor reward a microbial ingredient, although from 1 May 2028 ASC farm audits will need feed from an ASC-certified mill, and 15 Vietnamese mills are certified 33,47 . The revealed pull is that Vietnam already imports about 9,200 t of inactive yeast a year (HS 2102.20, food and feed together), 95% of it at about USD 1.1 per kg, and Vietnam already makes and exports the same product 32 . The largest global aquafeed group reached 0.6% novel ingredients against a 5 to 10% target 48 .

What changes in the ranking. With the two market-pull changes and the published weights, domestic textured plant protein (T1) leads the balanced preset at 3.95, against 3.9 for the cassava sugar hub (T4), and plant-drink proteins (T10) move up to third place for startups, joint third for manufacturers and fourth for investors. The lead is 0.05: a one-point fall in T1's market pull or defensibility returns first place to T4. T1 leads under the balanced, investor, startup and manufacturer presets; T4 stays first or second under every preset and leads under the policy, research and international presets, where the shared pilot and tolling play (T5) stays second. The frontier bets (T7, T8) stay where they were (our calculation from plays.csv, play_weight_presets.csv and play_demand_check.csv) .

Proposed supply-score changes, not applied. We have not changed plays.csv. The actor-check evidence suggests three supply-score changes, recorded in the note_v0_4 column of play_demand_check.csv for readers to apply with the interactive ranking in Plays: T1 capital efficiency from 4 to 5 and Vietnam advantage from 3 to 2, which nets to zero on the balanced score (T1 stays first at 3.95 and first under the investor, startup and manufacturer presets); and T2 defensibility from 3 to 2, on weak evidence, which would leave T5 sole third under the balanced preset (our calculation) .

Demand-side kill tests to add to the plays (play_demand_check.csv):

  • T1 fails if delivered food-grade defatted soy flour costs more than about USD 700 to 940 per t for a 1 t/h line at 70% utilisation (USD 590 to 900 for a 0.5 t/h line), if fewer than three buyers sign conditional orders totalling at least 2,000 t a year, or if India-origin textured soy (USD 0.98 per kg FOB) becomes the reference price (our calculation) 30,46 .
  • T2 fails if two head-to-head trials at 1 to 2% inclusion against the commodity yeast mills already buy show no gain worth the price gap (about 1.2 to 1.8% of shrimp-feed cost), or if the delivered price stays above about USD 1.1 per kg without such a gain 32 .
  • T7 is confined to export and business-to-business uses if purified precision-fermented proteins must carry a GMO declaration at the doses buyers use.
  • T9 fails if, at the replacement level that gives the same saving as soy, a fungal blend in industrial xúc xích or a coarse filling scores below the processor's own soy-extended formula in a blind test, or if no processor will pay above hydrated-soy parity (about USD 1.2 per kg dry at 45% protein) 27,10 .
  • T10 fails if an isolate cannot reach the band between Chinese and Western pea protein with UHT stability, or if host drinks cannot be made at or below 5 g of total sugar per 100 ml once plant milks are ruled inside the sugary-drink excise 17,19 .

25.4 Twenty-one cheap moves, and who would pay

Most of the demand gaps in Part III can be closed for tens of thousands of USD. The full list, with who, when and who would pay, is in demand_moves.csv; the underlying questions are in R1. Open questions. Every move now names a payer, or says why none was found. Only four of the original nineteen moves have a funder whose stated scope fits and that accepts Vietnam: two consumer-acceptance studies fit the Bezos Centre for Sustainable Protein at NUS, and two naming and GMO rules fit Nova Vista's market-shaping programme. Diet-change and culture-change work is excluded by Nova Vista and now by ACE Movement Grants, and ReRoot Asia excludes Vietnam 49,50,51,52 . 4 of 19 The rest belong to industry, to trade bodies or to national budgets.

Measure.

  • Chay days (DMV-01). First ask the Global Diet Quality Project or Gallup to tabulate the 2021 flesh-free share (7.9% of adults) by the lunar date of the recall day, then add a chay-day question to a probability survey 53,54. Who would pay: no philanthropic payer; national survey budgets (NSO, NIN) or GAIN's Diet Quality Questionnaire.
  • Customs declarations (DMV-02) for HS 3504, 2106.10, 1109 and 2102.20.10 (yeast for animal feeding), to split food from feed and name the buyers of T1 and T2 products. Who would pay: partly Nova Vista or World Bank and ADB technical assistance; otherwise a would-be T1 or T2 owner.
  • Sensory and choice tests (DMV-03) and a randomised label and name test (DMV-04). Who would pay: the Bezos Centre at NUS, with a Vietnamese institute.
  • One canteen and school trial with three arms (DMV-05): an extender or canteen ingredient in minced dishes; a chay-day menu; and a school arm in which one caterer that writes one menu for several schools serves a textured-soy or blended dish in the plant slot and in one minced dish, with daily published menus as controls. Log the dish species replaced, because animals spared per meal differ about 34-fold between a pork and a chicken dish and chicken leads only 13% of canteen protein dishes 55,56,57. Who would pay: the extender arm partly fits Nova Vista's procurement lever and the Bezos Centre; Humane World Viet Nam fits the chay-day arm as implementer, but no current funder both covers Vietnam and funds diet change; the school arm has no external payer.
  • A minimart repeat test (DMV-06), the share of meat eaten as processed products (DMV-09) and a Nha Trang hotel and menu audit (DMV-10). Who would pay: industry and AltProtein Vietnam; Humane World fits the hotel audit as implementer.
  • Quote requests to wrapped-food buyers (DMV-11): one vegetable and one plant-protein SKU from three EU importers of Asian frozen foods and two brand owners already sourcing from Vietnam, with their certification and origin requirements. Retailer protein targets need no survey; they are published 58,59. Who would pay: industry and trade bodies.
  • A soy-flour price sheet (DMV-21): five delivered quotes for food-grade defatted soy flour and three for Chinese plain textured soy, with the T1 breakeven. It costs tens of hours and decides the T1 kill test 30,46. Who would pay: the would-be owner or its lender.

Build public goods.

  • A chay safety and composition survey (DMV-07) that tests borax, preservatives, microbes and protein first, with animal DNA as a cheap PCR screen; publish with VFA, publish methods and do not name small makers 7. Who would pay: partly the Bezos Centre, UNICEF Viet Nam or ACIAR; otherwise food-safety budgets.
  • A standing panel of research and procurement buyers (DMV-08) to publish specifications, minimum orders and qualification times. Who would pay: an industry association.
  • A feed carbon-footprint benchmark (DMV-12) that helps ASC-certified mills meet GHG reporting; it will not by itself create demand for microbial ingredients 33. Who would pay: partly ADB Ventures, IFC or ACIAR; otherwise the mills.

Set rules (policy makers). These add a demand-side reason to the policy options of Policy options. The Food Safety Law takes first comments at the National Assembly's October 2026 session and is considered for passage at the May 2027 session, so a new-food clause needs to be in the draft revised between the two sessions, in practice by early 2027 60 . Two of the four label moves are largely law already and need enforcement and testing, not new rules; one should defend a permissive regime; one shrinks to a single question 24,25,23 . Decree 37/2026 has replaced Decree 43/2017 for goods labels, and Circular 30/2026/TT-BYT has replaced Circular 29/2023 for nutrition labels.

  • Chay claims (DMV-13). TCVN 13625:2023 already defines vegetarian and vegan claims, but its vegetarian class allows egg, milk and honey and it says nothing on alliums. Ask VFA to state that an unqualified chay or thuần chay (strict vegetarian) claim means the vegan criteria, with egg or milk named in a qualifier, and enforce the existing rules alongside DMV-07.
  • Blend labels (DMV-14). Any meat named in a product name must already be quantified. Narrow the move to a voluntary industry code for generic names such as giò lụa (steamed pork roll) and xúc xích that states the meat share when a plant or fungal protein replaces 10% or more of the meat; its value is defensive, as proof against a độn accusation 24,11. Who would pay: the processors.
  • Names (DMV-15). No rule reserves thịt (meat), sữa (milk) or phô mai (cheese); names must only not mislead, and practice already qualifies meat words. Defend this permissive regime in the new Food Safety Law decree and aim vocabulary guidance at state and media use.
  • GMO status (DMV-16). The GMO label applies only above 5% of ingredients, so lactoferrin-type uses (about 0.08%) are not caught. Ask MAE and VFA one question: is a recombinant protein identical to the native protein a detectable product of the modified gene, and may a product containing one below 5% still claim non-GMO? Who would pay for DMV-15 and DMV-16: Nova Vista, or for a small Vietnamese organisation ACE Movement Grants (policy work only, deadline 18 October 2026).
  • School menus (DMV-17). National guidance already has no animal share. Ask MOET, NIN and Ajinomoto to replace the menu software's 48% animal-protein default with protein-quality, zinc, iron and sodium targets per lunch, and to add textured soy and legume products to the software's food list, inside the MOET circular due in 2026 and the digital tools due in 2027 to 2028 1,15. A second version is a procurement standard for innovative protein in public and state-enterprise canteens. Who would pay: UNICEF Viet Nam's 2027 to 2031 programme for the school version; Nova Vista's procurement lever for the canteen version.
  • Customs subheadings (DMV-18) for soy isolate, pea protein and textured protein. Who would pay: MOF's own budget, partly World Bank or ADB.
  • An origin note for exporters (DMV-19) on the EU cap of 20% non-originating flour and starch in wrapped foods and the 10% tolerance for imported textured soy in 2106 foods, and on claiming CPTPP for the UK; imported soy protein is not the main constraint 31. Who would pay: industry and trade bodies.
  • A written sugary-drink ruling (DMV-20): the Ministry of Finance to state, before 1 January 2027, whether soy milk (TCVN 12443:2018), nut and cereal milks above 5 g of total sugar per 100 ml are nước giải khát (soft drinks) 16,61. Who would pay: the plant-milk makers.

25.5 What not to do

  • Do not size a business on stated survey intent or on the published Vietnamese market sizes; neither has a revealed basis.
  • Do not launch premium imported-style analogues into gourmet channels expecting mass uptake, and do not enter foodservice with an imported branded analogue through business-to-business distributors: CP Meat Zero minced was listed on Kamereo at 1.5 to 2.1 times CP's own minced pork per kg, and none of its eight SKUs was orderable in September 2026 62,12 .
  • Do not price a mass product above pork, a chay product above today's chay shelf, or an extender above hydrated soy without proof that it allows more undetected replacement.
  • Do not lead with "meat replacement", "the future of meat", "lab" or environmental claims, and do not use độn or "x% meat replaced" wording; lead with taste in familiar dishes, protein, safety and origin, and name the plant part as an ingredient.
  • Do not name products thịt nhân tạo (artificial meat) or market cultivated or precision-fermented products to chay buyers.
  • Do not start in schools with a branded or meat-replacement product. The school plant slot is small (about 1,000 t of protein a year, our calculation), filled by tofu and under-used; enter later as a caterer ingredient that beats tofu on cost, or as a labelled blend in minced dishes 14,13 . Do not start in wet markets or tourist hotels either.
  • Do not pitch a fortified soy dish to upland boarding or border-commune kitchens as filling a meatless lunch: their published lunches carry meat almost every day, and a Decree 339 lunch spent 91 to 97% of its non-rice money on pork. Pitch cost per gram of protein against pork, which these kitchens already stretch with tofu 2,28 .
  • Do not present import substitution, or protein upgrades on existing chay days, to welfare or climate funders as meat displaced, animals spared or CO2e avoided; they deliver none 52,57 .
  • Do not count sales into existing chay days as meat displaced.

25.6 What would change our view

  • A tested Vietnamese study (choice experiment or in-store trial) showing repeat purchase of a modern analogue or blend at 20% or more above the meat it replaces.
  • Retail or e-commerce panel data showing sustained growth of modern plant-based meat since 2024.
  • A probability survey, or a lunar-date tabulation of the 2021 diet survey, showing that chay keeping is rising among young urban people, which would make added chay days a real displacement route.
  • A large processor launching a declared blend with a novel protein, or a formula maker sourcing a precision-fermented protein.
  • A new Food Safety Law, now expected to be considered for passage in May 2027, with a working new-food route and a clear treatment of purified fermentation proteins.
  • Delivered soy-flour quotes below about USD 700 per t, which would make a domestic textured-soy line pay even at the high-cost end.
  • A food-safety shock in meat followed by a measurable switch to plant proteins, which the 2019 to 2020 pork shock did not produce.

For each reader.

  • Startups: sell a familiar food with little capital, and aim first for ingredient, private-label or co-packing supply to incumbents. A restaurant menu listing for your own analogue brand is not a business customer: in the region such firms mostly went quiet or pivoted. Of 24 regional and Vietnamese firms since 2019, 7 are active, 4 pivoted, 2 exited and 11 are dormant; none of the 4 Vietnamese meat-analogue startups is active, while all 3 makers of traditional formats are. No founder has yet tested the ingredient route, so treat it as the hypothesis to test, not a proven path 63,64,65 .
  • Investors: T1 is a debt-financed import-substitution plant for a chay maker, co-packer or distributor, not a venture thesis; underwrite the delivered soy-flour price, conditional orders and cost per gram of protein. Investors now say they fund offtake, contracted revenue and unit economics, which T1 and T10 can show and consumer analogues cannot 66,67 .
  • Manufacturers: a domestic textured soy, a non-soy isolate for drinks at or below 5 g of sugar per 100 ml, and a fungal ingredient that beats your own soy-extended formula are the three new ingredients your products can use. Test blends first in industrial xúc xích and coarse fillings, not giò lụa, blind and at equal saving against your soy recipe; keep 10 to 20% replacement in fine emulsions and 30% in coarse mince; and consider private-label chay through an existing co-packer 9,20 .
  • Research bodies: the tested-evidence gap (sensory, choice and label tests in Vietnamese formats) is the most useful thing you can close for the whole sector, and the Bezos Centre at NUS is the one funder whose scope fits it.
  • Policy makers: most label rules already exist; enforce them, defend permissive naming, answer the single GMO question, rule on plant milks and the sugary-drink excise before 2027, and put protein quality, zinc and cost into the school menu software.
  • International organisations and funders: match the route to your unit and your rules. Welfare funders: a canteen trial with Humane World Viet Nam, the one implementer with Vietnam staff, focused on chicken dishes, which lead only one protein dish in eight. Market-shaping funders: canteen procurement and the naming and GMO rules; Nova Vista's programme funds procurement, advance market commitments and retailer ratio targets, excludes whole-food, culture-change and diet-change work, and prioritises Japan and Korea. R&D funders: isolate and fungal functionality and Vietnamese blind tests. Nutrition agencies: protein quality and zinc in school food standards. On the benchmark path the displacing routes spare about 2.4 to 2.7 million animals a year by 2035, 93 to 94% of them chickens and ducks, and avoid about 0.14 Mt CO2e, so climate is a co-benefit, not the case; corporate pledges are a weak lever, since only Accor's dish target binds Vietnamese units 49,68,57,69 . The chay-day survey item belongs in national surveys.

Related: Plays, Policy options, Demand sizing, Actor check, M4. Actor check and waves, R1. Open questions.

Nguồn được trích dẫn trong trang

Toàn bộ nguồn (Phụ lục T)

  1. PMR-26 Nguyen Trai Primary School, Hanoi; Ngoi Sao Xanh food company (Bac Ha group). Thông tin dinh dưỡng của thực đơn. 2026-03 to 2026-05
  2. UPL-01 Muong Nhe PTDTBT lower secondary school (Dien Bien). Thực đơn ăn bán trú tuần 3 tháng 9/2026. 2026-09-18
  3. UPL-09 Leng Su Sin PTDTBT lower secondary school (Sin Thau, Dien Bien). Thực đơn tuần 26. 2026-03-02
  4. APR-27 UNICEF. Viet Nam 2027-2031 Country Programme Document: For Every Child, Every Right. 2026-07-06
  5. NQR-01 National Institute of Nutrition. Tra cứu nhu cầu dinh dưỡng. 2026 (age groups updated 22 to 23 Mar 2026)
  6. MKT-03 Lazada. Lazada Việt Nam search results. 2026-09-25
  7. TRU-24 Quality Assurance and Testing Centre 3 (QUATEST 3), state technical centre. QUATEST 3 triển khai phương pháp phát hiện DNA động vật trong sản phẩm hoặc nguyên liệu thực phẩm và thức ăn chăn nuôi. 2025-02-04
  8. BUY-05 VISSAN. Vissanmart processed food product pages. accessed 2026-09-24
  9. HXE-01 Ho Chi Minh City University of Technology (VNU-HCM). Identifying the drivers of liking for texture attributes of Vietnamese sausage. 2021-05-10
  10. TRU-02 Trung tâm Kiểm soát bệnh tật tỉnh Quảng Ninh (provincial CDC). Cách nhận biết giò chả an toàn ngày Tết. 2015-02-09
  11. TRU-03 Lao Động. Giò chả "tự cung" để đối phó giá thịt lợn quá cao. 2019-12-18
  12. CPC-12 Kamereo. Kamereo product pages. 2026
  13. AIB-39 Ministry of Health, Viet Nam. Quyết định ban hành Hướng dẫn dinh dưỡng đối với bữa ăn học đường. 2025-12-25
  14. PMR-10 Nguyen Binh Khiem Primary School, Sai Gon ward, Ho Chi Minh City. Thực đơn bán trú năm học 2026-2027. 2026-09
  15. PMR-01 Prime Minister of Viet Nam. Quyết định 973/QĐ-TTg phê duyệt Chương trình Sức khỏe học đường giai đoạn 2026-2035. 2026-06-01
  16. FTR-01 National Assembly of Viet Nam. Luật Thuế tiêu thụ đặc biệt 2025 số 66/2025/QH15. 2025-06-14
  17. FTR-02 Government of Viet Nam (signed by Deputy PM Ho Duc Phoc). Nghị định 360/2025/NĐ-CP. 2025-12-31
  18. FTR-05 Ministry of Science and Technology; drafted by TCVN/TC/F9 Đồ uống. TCVN 12828:2019 Nước giải khát. 2019
  19. FTR-32 Ministry of Industry and Trade. Vietnam National Trade Repository: tariff schedules. accessed 2026-09-25
  20. CPC-15 Saigon Co.op (Co.op Online). Co.op Select product pages. 2026-09
  21. AIB-28 Thanh Niên. Saigon Co.op thêm giá trị cho giỏ hàng Việt từ hàng nhãn riêng. 2026-09-25
  22. LBL-03 MOST; drafted by TCVN/TC/F6, proposed by the Vietnam Standards and Quality Institute. TCVN 14429:2025 Hướng dẫn sử dụng công bố dinh dưỡng và công bố khuyến cáo về sức khỏe. 2025
  23. LBL-15 MOST (Decision 586/QĐ-BKHCN, 2023); drafted by TCVN/TC/F3. TCVN 13625:2023 (ISO 23662:2021). 2023
  24. LBL-06 Government of Viet Nam. Nghị định 43/2017/NĐ-CP về nhãn hàng hóa. 2017-04-14
  25. LBL-10 MARD (now MAE) and MOST. Thông tư liên tịch 45/2015/TTLT-BNNPTNT-BKHCN. 2015-11-23
  26. DIE-25 FAO. FAOSTAT Producer Prices. 2026
  27. HXE-10 US Department of Agriculture (eCFR). 7 CFR Part 210 Appendix A (Alternate protein products). current as accessed
  28. UPL-04 Muong Nhe PTDTBT lower secondary school (Dien Bien). Công khai thực đơn ăn trưa theo Nghị định 339/2025 tháng 9 năm 2026. 2026-09-18
  29. BUY-02 United Nations Statistics Division (UN Comtrade). Exports to Vietnam of protein ingredients by reporter, 2019 to 2025 (mirror data). 2019 to 2025 data; queried 2026-09-24
  30. TIC-12 United Nations Statistics Division (UN Comtrade). China exports to Vietnam of soy flour, soybean meal, textured protein, gluten and other proteins, 2023 to 2025. 2026
  31. ORG-01 European Union; Socialist Republic of Viet Nam. EU-Viet Nam Free Trade Agreement, Protocol 1 and Annexes I and II. 2020-06-12
  32. FBA-17 United Nations Statistics Division. Mirror imports of HS 210220 into Viet Nam. 2026-09-25
  33. FBA-01 Aquaculture Stewardship Council. ASC Feed Standard v1.2. 2025-12-19
  34. EXP-29 Aquaculture Stewardship Council. ASC Feed Standard v1.1. 2025-05-01
  35. EXP-39 United Nations Statistics Division. UN Comtrade: Vietnam exports by HS 6-digit code, 2023. 2026-09-24
  36. NOV-59 VnExpress (sponsored by VitaDairy). VitaDairy giới thiệu ColosBaby Lactoferrin 3+ với công thức miễn dịch đa thành phần. 2026-05-07
  37. NOV-10 Green Queen. Good Meat: Eat Just Debuts Cultivated Chicken in Singapore Retail. 2024-05-16
  38. AIS-34 Protein Production Technology International (PPTI). Singapore cell-cultivated seafood firm Avant Proteins winds up local operations. 2026-02
  39. AIS-35 PetfoodIndustry. Umami Bioworks, Friends & Family Pet Food launch cultivated seafood for pets. 2025-09-03
  40. NOV-63 VnExpress (sponsored by Vinasoy). Lý do Veyo sữa hạt lên men tự nhiên thu hút người dùng. 2024-10-11
  41. BUY-16 Vinamilk. Sua hat Cao dam it duong (product page). accessed 2026-09-24
  42. TIC-14 United Nations Statistics Division (UN Comtrade). China exports of textured protein and wheat gluten by partner, 2025. 2026
  43. TIC-02 Jinan Sunward Machinery Co., Ltd.. Turnkey Twin-Screw Textured Soy Protein Tsp Tvp Project Processing Line. 2026
  44. TIC-03 Jinan Saibainuo Extrusion Machinery Co., Ltd.. Tvp Tsp Fiber Artificial Vegetable Meat Production Plant. 2026
  45. TIC-07 Shandong Joyang Machinery Co., Ltd.. Tvp Tsp Textured Vegetable Soy Protein Soya Chunks Nuggets Minces Making Machine. 2026
  46. TIC-13 United Nations Statistics Division (UN Comtrade). Soya bean flour and meal exports by main exporters, 2024 and 2025. 2026
  47. FBA-04 Aquaculture Stewardship Council. ASC Find a Feed Mill. 2026-09-25
  48. FBA-07 Nutreco. Nutreco Impact Report 2025. 2026
  49. AFN-05 Nova Vista Foundation. RFP: Protein Innovation. 2026
  50. AFN-15 Bezos Earth Fund. Bezos Centre for Sustainable Protein Opens at the National University of Singapore. 2024-09-05
  51. AFN-29 Animal Alliance Asia. ReRoot Asia Fund. 2026
  52. FUF-01 Animal Charity Evaluators. Movement Grants: Call for Applications and Introducing Our New Grants Manager. 2026-09-24
  53. APR-36 Global Diet Quality Project. Viet Nam Data: Global Diet Quality Project. 2024
  54. BRD-03 Gallup; Global Diet Quality Project. Country Data Set Details DQQ: Gallup Worldwide Research data collected 2021-2024. 2024
  55. UPL-27 Hoang Kim industrial catering (Binh Duong). Thực đơn: Suất ăn 20.000, 23.000, 25.000. 2024-02-25
  56. UPL-32 Haseca. Thực đơn suất ăn công nghiệp theo tuần, tháng. 2025-12-11
  57. FUF-03 Food and Agriculture Organization of the United Nations. Production: Crops and livestock products (bulk download, Asia). 2025-12-23
  58. XBA-07 Lidl (international purchasing). Supplier list, Food (2025 deliveries). 2026
  59. XBA-20 CJ CheilJedang. CJ제일제당 베트남 키즈나 공장 준공. 2022-02-28
  60. APR-07 Dan Tri. Chủ tịch Quốc hội: Sửa luật, thực phẩm đến tay người dân có an toàn không?. 2026-09-24
  61. FTR-13 Thư Viện Pháp Luật. Công văn search results. 2026-09-25
  62. AIS-25 Kamereo. Heo Xay Chay Meat Zero 1kg. 2026
  63. AIS-28 TiNDLE Foods via PR Newswire. TiNDLE Foods Pivots Business Model to Focus on Private Label Products. 2025-11-20
  64. AIS-39 We-Fi. Pioneering Sustainable Agri-Tech and Plant-Based Proteins from Vietnam. 2025-12-15
  65. BRD-08 Green Queen Media. Karana Announces Global Expansion for Sustainable Jackfruit Meat. 2023-04-18
  66. AIS-01 FoodNavigator. Alternative protein funding diverges as investors prioritize scale and bankability. 2026-02-18
  67. AIS-15 Vietnam Private Capital Agency (VPCA) with Do Ventures and NIC. Vietnam Innovation & Private Capital Report 2026. 2026
  68. AFN-23 Humane World for Animals. Giving thanks for progress for farmed animals all over the world. 2025-11-26
  69. AFN-40 Accor. Accor Group Good Food Policy. 2024-11