April 15, 2026 ·26 min
On Competition
Pinduoduo's competitive landscape — Douyin and Alibaba at home; Amazon, TikTok Shop, and MercadoLibre abroad.
Contents 5 sections
PDD’s success cannot be overstated. As a company, it went from founding to surpassing Alibaba in cash reserves in under a decade. The horizon has now changed. The hunter is starting to become the hunted.
PDD faces stern competition from China (Douyin + Baba + JD) and internationally (primarily Amazon + TikTok + MercadoLibre + AliExpress). The extraordinary cash position ($60.4B) provides strategic optionality.
Douyin / TikTok Shop, in particular, is a dangerous source of competition. Its funnel is unique and leverages a unique flywheel. To visualize its future impact, I must borrow a concept learned from Mr Li Lu: 势不可挡 — like the momentum of a sequoia seed that has the biological inevitability to become a behemoth.
PDD is getting serious. The co-CEO structure has Chen Lei overseeing Temu and Zhao Jiazhen managing domestic Pinduoduo. Neither executive seems to be siloed. PDD covers 90+ countries with meaningful scale. But, as confirmed in the previous earnings call, corporate governance and internal talent lag behind business growth amidst a complicated geopolitical environment.
Temu’s shift away from a fully managed model impacted margins but was necessary for scalability. Controlling pricing, logistics, marketing, and customer service for millions of SKUs across 90+ countries would bring an operational complexity that grows geometrically with scale.
At the peak of the fully managed model, Temu incorporated elements of disruptive innovation and subsidized arbitrage. It served overlooked customers — value-conscious, willing to accept long delivery times for dramatically lower prices — and built a market previous firms (AliExpress et al.) never truly cracked: ultra-low-price cross-border e-commerce at scale. It subsidized pricing below cost through mass marketing, operating at a loss to gain share. It was very successful at this.
Each platform’s recommendation engine runs a multi-objective optimization balancing engagement, revenue, and retention (or acquisition, for new-customer interactions). Each competitor weights these differently. Let’s assess the field under Porter’s lens of competitive rivalry.
Pinduoduo (China)
China remains the world’s largest e-commerce market (2024: $2.16T in online retail sales). Continuing to succeed here is imperative for Pinduoduo.
In Nov-2025, China’s SAMR released draft platform antimonopoly guidelines. In Feb-2026, the finalized rules largely banned exclusivity clauses — allowing multi-platform merchants to proliferate — and banned using algorithms to coordinate pricing or traffic distribution with competitors.
In Dec-2025, Pinduoduo MAU was 720MM (+1% YoY) and DAU was 395MM (+4% YoY), per QuestMobile. The last official MAU reporting by Pinduoduo was 751.3MM in Q1-2022. Back in Jul-2022, QuestMobile showed Pinduoduo with a DAU/MAU ratio of 60.4% (+5.5pp YoY), versus Taobao’s 40.4% and JD.com’s 21.3% — proof of higher engagement, driven by gamification (virtual farming games, daily check-in rewards).
Pinduoduo has an entrenched value positioning in China. Its profits continue to subsidize Temu’s international expansion.
Selling fees
Pinduoduo’s selling fees are low.
| Fee type | Official name | Applies to | Rate |
|---|---|---|---|
| Basic Technical Service Fee | 基础技术服务费 | All orders, all merchants (was non-refundable on refunded orders until Sep-2024) | Flat 0.6% of (payment + platform coupon) |
| Technical Service Fee (by category) | 类目技术服务费 | Specific categories only (virtual goods, jewelry, pharma, etc.) | 1%–5%+ by sub-category |
| ”Billion Subsidy” Promo Fee | 百亿补贴技术服务费 | Orders from Billion Subsidy campaigns | 1.4%–4.4% by category |
| Livestream Technical Service Fee | 直播技术服务费 | Orders completed via livestream | 0.4%–3.4% for most categories |
Pinduoduo applies a base + highest-applicable-fee rule. Every order pays the 0.6% Basic Technical Service Fee. When an order triggers multiple fee types at once (e.g. a jewelry item sold via Billion Subsidy livestream), Pinduoduo charges only the highest single applicable fee in addition to the base. Fees do not stack (择高收取，不叠加).
For Billion Subsidy fees, Pinduoduo has raised rates on mature categories while lowering them on industrial/growth categories.
| Category | Fee | Recent change |
|---|---|---|
| Smart electronic locks | 4.4% | New in Jan-2025 |
| TCM health instruments | 4.4% | Up from 2.4% (Sep-2025) |
| Oral care (water flossers, irrigators) | 4.4% | Up from 1.9% (Sep-2025) |
| Motorcycle rider equipment | 4.4% | Up from 1.4% (Nov-2025) |
| Hair masks / conditioners, soap | 3.4% | Up from 1.9% (May-2025) |
| Auto care (chassis armor, disinfectants) | 2.4% | Up from 1.4% |
| Industrial pipe fittings / standard parts | 1.4% | Down from 4.4% (Apr-2025) |
| Most standard physical goods | 1.4%–3.4% | Typical range |
The fee stack means Pinduoduo’s lower-bound selling-fee take rate on GMV is 0.6% while the upper bound hovers ~5%. Ad revenue augments this take rate.
Continuing key strengths
The ability to serve a high volume of goods (underpriced vs competition) with specialized manufacturer-direct sourcing is a quality similar to Costco. This model creates a flywheel: low prices → high volume → supplier dependence → lower prices.
The reverse-auction system — merchants bid down prices for traffic exposure — is an algorithmic market design no competitor has replicated at equivalent scale. Under Porter’s five forces, this mechanism suppresses supplier power.
Pinduoduo also helps willing sellers improve product quality through industry insights and supply-chain support, and is building infrastructure (transfer warehouses, last-mile pickups) to overcome high last-mile logistics cost in remote rural areas. Its agricultural focus is aligned with CCP priorities.
Douyin’s advantage lies in its attention monopoly and content-driven discovery. Pinduoduo’s lies in pure price competitiveness and established purchasing habits among value-conscious consumers.
Douyin
Douyin built its e-commerce business (from 2018) on livestream shopping — content-based commerce, discovering products while scrolling videos. (For context: total live commerce generated $807B in GMV in 2024, per the International Trade Administration.) It has had a standalone Mall app since mid-2022 (400MM+ Android downloads), and ByteDance launched Douyin Pay in 2021.
ByteDance’s ability to cross-subsidize commerce with ad revenue and content engagement creates a cost structure difficult for Pinduoduo to match on a unit-economics basis. It leverages its entertainment-driven platform to reduce customer-acquisition and retention costs — benefiting both COGS and S&M margins.
Douyin is now shifting toward shelf-based commerce (search bars, product listings, a mall tab) — directly attacking Pinduoduo’s core model. This makes sense as Douyin scales: livestream commerce is unpredictable, dependent on individual influencers, hard for merchants to plan around, and expensive. In 2022, President Wei Wenwen set a goal of shelf-based commerce exceeding 50% of sales, on a three-year plan (30% of GMV in 2023, 40% in 2024, 50% in 2025).
In the 12 months from Aug-2024 to Jul-2025, Douyin e-commerce GMV grew 34% YoY; shelf-based GMV was +49%. Douyin saw RMB 3.5 trillion ($483B) in GMV in 2024 (+30% YoY).
Douyin is capturing a disproportionate share of incremental GMV growth, penetrating Pinduoduo’s core lower-tier-city base. Most external reports cite 60%+ of new livestream buyers coming from tier-3 and smaller cities — historically Pinduoduo’s core territory. Douyin is growing the pie and likely acquiring existing Pinduoduo users too.
I cannot quantify which % of Pinduoduo’s user base is at risk. This is important but currently unknowable without granular data. Anything else would be anecdotal, extrapolated, or simply fooling myself.
Douyin’s merchant ecosystem is expanding rapidly: 162,000+ shops exceeded RMB 1MM in livestream sales by mid-2025 (+113% YoY). In Jun-2025 it eased onboarding — new merchants can start with zero deposit, paying gradually after 200 orders or RMB 10k in sales; base deposits were cut ~75%. In Mar-2025, Ocean Engine (巨量千川, Douyin’s ad division) was merged into Douyin e-commerce, better aligning its ad and organic traffic pools.
A KPI to watch is merchants shifting ad budget from Pinduoduo to Douyin. I’ve seen some merchant-survey data confirming this (but such surveys are rarely without bias).
Taobao + Tmall
The involvement of Tencent’s WeChat helped Pinduoduo proliferate against these incumbents. Tencent still owns ~14% of PDD (754.36MM Class A shares) per the 2024 20-F.
The Taobao + Tmall combo remains China’s biggest B2C marketplace since launching in 2008. Alibaba last officially reported “China commerce” MAU at Q1-22 end: 903MM annual active consumers, of whom 124MM each spent more than RMB 10,000 in FY22.
Alibaba has undergone multiple restructurings — from “1+6+N” (Mar-2023) to a 4-group structure (Aug-2025). The recent success of the Qwen models brings positive sentiment that masks governance cracks. Its reactive responses to Pinduoduo (price-matching subsidies, merchant incentives, UX investments) have compressed margins: Taobao & Tmall Group EBITA grew only +1% in FY25.
JD
CEO Sandy Xu confirmed annual active customers surpassed 700MM in Nov-2025 — though continued frequency growth was driven by JD Food Delivery. JD continues to occupy its position through a logistics-first model (listed JD Logistics): 3,600+ warehouses spanning more than 34MM m², plus 130 overseas warehouses in 23 countries, and the new JoyExpress last-mile brand.
Temu (US)
Temu looks very different from the fully managed model launched in Sep-2022. The original approach consolidated products at Chinese warehouses and air-freighted directly to consumers; Temu controlled listing, pricing, marketing, fulfillment, and customer service. It selectively used foreign warehouses for market-tested bestsellers unlikely to cause inventory backlogs.
Since May-2024, the transition to a semi-managed model (merchants with overseas warehouses handle logistics; Temu retains pricing + platform control) has been ongoing. Semi-managed products were priced 10%–20% below Amazon; fully managed goods carried deeper discounts (more than 50%). In Nov-2024, Temu allowed local US sellers to onboard.
The elimination of the US de minimis exemption (China: May-2025; global: Aug-2025; statutory by Jul-2027 via OBBBA) forced Temu to abandon its direct-from-China US model. In Apr-2025 it launched the Y2 semi-managed model (TEMU 半托 Y2 模式): instead of pre-stocking, the seller ships each order from China via air freight after it’s placed, to a US warehouse where the delivery label is changed. Sellers cannot mark “shipped” until goods actually reach the US.
| Feature | Standard semi-managed | Y2 semi-managed (Apr 2025) |
|---|---|---|
| Pre-stocked in US warehouses | Yes | Ships direct from China per order |
| Warehouse storage fees | High (bulk pre-positioning) | None |
| Upfront capital | High | Low (ship as orders come) |
| Fulfillment timeline | 1–5 days (already in US) | Up to 14 working days (9 prep + 5 delivery) |
| Shipping flow | US warehouse → customer | China → air freight → US customs → last-mile |
| Last-mile carriers | Standard platform logistics | USPS, FedEx, UPS via platform portal |
| ”Shipped” button rule | Mark when dispatched | Cannot mark until goods reach US |
| Customs | Seller or agent handles | Platform offers full customs + tax-inclusive service |
The duty-free, air-freighted, direct-from-China parcel arbitrage is gone. Every path forward requires tariff absorption, local inventory, or margin compression. All cross-border marketplaces face the same tariff regime but have very different structural resilience — Temu is extremely exposed, as its value proposition was most dependent on price arbitrage. It has prepared a B2B2C customs-clearance model: this solves the regulatory problem but adds a cost layer that must be absorbed by the customer or PDD’s margins. Neither is good, and I don’t see an alternative at this time.
Ex-tariff, cross-border shipping from China to the US was ~$3–8 per small parcel with zero duty. Local fulfillment costs $3–7 per order but adds US warehousing ($0.8–$2.5/ft³/month). The expected ROI on local warehousing is unknown.
I want the platform to be sustainable for the long term. I want Temu to remain laser-focused on curating the best user experience with light monetization.
US warehouses
Cross-border tariffs into the US are deeply volatile, necessitating warehouse presence. Post-tariff (at ~17.5%–35% Section 301 + Section 122 rates), formal customs entry adds $1–5 in brokerage plus tariff. At 25% duty on a $40 order, that adds $10 in tariff alone; at the peak 120% rate, the same order added $48 — exceeding the goods value.
As of 2025 end, Temu has at least 13 self-owned warehouses globally, including 3 in the US and one near the US–Mexico border. It ships via ocean freight to hubs in Rowland Heights CA, Dallas TX, and Newark NJ, plus a 120k m² facility at the border. Items ship in 1–3 days with “Local Warehouse” badges and no import charges.
| Hub | Location | Operator | Coverage |
|---|---|---|---|
| West Coast | Rowland Heights / City of Industry, CA | WINIT (万邑通) America | CA and nearby states |
| Central | Dallas–Fort Worth, TX | Various 3PL partners | TX, OK, LA, AR, NM, central |
| East Coast | Newark, NJ | Easy Export (出口易) + regional 3PL | NY, NJ, PA, CT, MA, Northeast |
| US–Mexico border | Near San Diego / Otay Mesa | JUSDA (Jitu Warehouse) | 120k m² returns + resale hub |
Temu sends returned goods to its US–Mexico Jitu warehouses, where they are repackaged and resold to reduce losses. In addition to self-operated sites, it has nearly 20 officially certified service warehouses in the US. In Dec-2025, Temu launched an official integration app on the Shopify App Store (2.25MM SMB merchants), providing one-click product syncing, real-time inventory, and order management across 30+ markets.
The true long-term threat to the supply chain would be total automated US-based manufacturing. I’d still put the probability of this existing within 5 years at 10%, and within 10 years at markedly higher.
Managing fakes
Temu has been very good at cracking down on fakes — I can’t find brand replicas on it, whereas they are openly available on AliExpress and DHGate. In May-2025, Temu signed an MoU with the International Anti-Counterfeiting Coalition and joined as a member (alongside Amazon, eBay, Apple, Nike, Chanel). Its measures include 24/7 algorithmic monitoring, a dedicated IP-protection portal and brand registry, mandatory seller compliance training, and an internal enforcement team.
TikTok Shop
The Jan-2026 resolution of the TikTok US ownership dispute (Oracle, Silver Lake, and MGX taking 15% each; other investors 5%; existing investors 30.1%; ByteDance retaining 19.9%) removes the existential regulatory risk and lets TikTok Shop invest aggressively. It remains the most effective attempt to blend content with commerce.
Creator-driven discovery is a differentiated value proposition neither Amazon nor Temu can replicate. A 6% commission rate (about half Amazon’s effective rate) and the mandatory Fulfilled by TikTok infrastructure help ecosystem development — before even accounting for TikTok’s superior affiliate engagement (more than 10% for smaller creators, vs under 5% on Instagram).
TikTok’s two-tier collaboration model enables sophisticated merchant–creator matching. Open Collab makes products visible to all eligible creators, maximizing reach. Target Collab is invite-only, letting brands hand-pick creators at negotiated commissions (typically 18–25% for proven performers vs 10–12% on open plans). The most effective sellers run both simultaneously. (TikTok also offers rare Exclusive Collabs, where the platform curates seller–creator matches and co-funds incentives.)
Unlike Amazon, TikTok’s model generates organic content that keeps driving discovery long after the initial commission is paid — a long-tail effect that sustains the flywheel.
On brand-willingness to adopt the platform — which Temu and AliExpress do not possess — here are two TikTok Shop execs I found worth quoting:
I actually think if I asked my daughter (17) what Dove stood for, I don’t think as a consumer she would be able to say “oh, they’re about body inclusivity.” I think it’s an even playing field, whether you’re an older, traditional CPG brand or a challenger just coming into the space. You have to constantly communicate to consumers what you stand for — and it means more to Gen Z than anybody else.
If you’re a brand with legacy and brand equity, you should leverage that in how you show up to TikTok. There’s nostalgia and vintage culture that are big trends — peering back into your marketing treasure chest. A number of brands have had success bringing back old jingles and reinvigorating them. Be smart about how you use your brand traditions and equity.
Fulfilled by TikTok (FBT)
FBT launched in the US, operating from warehouses in Pennsylvania and Virginia; products show a “Free 3-Day Delivery” badge. It’s similar to FBA — sellers pay per-unit fulfillment fees by weight, storage fees by duration, and potential hub-placement fees. Over the last year, TikTok discontinued split inbound shipments, raised hub-placement fees, removed the $30 free-shipping threshold (all FBT products now ship free by default, with an opt-out for low-margin sellers), and eliminated seller-funded free shipping. In Dec-2025 it made FBT storage much cheaper (−14% to −43% for inventory up to 270 days, first 60 days free). From Dec 15 2025, cross-border sellers must use TikTok Logistics exclusively; new US sellers from Feb 9 2026. On Jan 12 2026, multi-unit FBT fees were cut up to −24% in the 0–4 lb range, incentivizing multi-item orders.
Amazon
Amazon has allowed 3P sellers since 2000, launched Prime in 2005, and FBA in 2006. Its marketplace take rate is 10%–30% (referral + FBA + advertising). The FBA network provides a service level cross-border platforms cannot match: as of Sep-2025, 5B+ items are processed annually through Amazon Logistics (~1MM robots). In 2025, Prime members received 13B+ items worldwide same/next-day; in the US, 8B+ within a day (+30% YoY).
Amazon’s $68.6B advertising business (+22% YoY; Q4-25 +23%) is a self-reinforcing flywheel: sellers must buy Amazon Ads for visibility, generating high-margin revenue while raising switching costs.
| Shipping weight | Low (<$10) | Standard ($10–$50) | Premium (>$50) |
|---|---|---|---|
| ≤ 4 oz | $2.91 | $3.68 → $3.73 | $3.68 → $3.99 |
| 4–8 oz | $3.13 | $3.90 → $3.95 | $3.90 → $4.21 |
| 8–12 oz | $3.38 | $4.15 → $4.20 | $4.15 → $4.46 |
| 12–16 oz | $3.78 | $4.55 → $4.60 | $4.55 → $4.86 |
| 1.0–1.25 lb | $4.22 | $4.99 → $5.04 | $4.99 → $5.30 |
| 1.25–1.5 lb | $4.60 | $5.37 → $5.42 | $5.37 → $5.68 |
| 1.5–1.75 lb | $4.75 | $5.52 → $5.57 | $5.52 → $5.83 |
| 1.75–2.0 lb | $5.00 | $5.77 → $5.82 | $5.77 → $6.08 |
| 2.0–3.0 lb | Flat hold on 2025 base | +$0.05 on 2025 base | +$0.31 on 2025 base |
| 3.0–20 lb | $6.15 +$0.08/4oz | $6.97 +$0.08/4oz | $7.23 +$0.08/4oz |
Amazon froze FBA rates for the <$10 tier across all standard weights in 2026 — directly competing with Temu’s core SKU price range. It now operates 10 regional US clusters (up from 8), and its “Add to Delivery” feature (Oct-2025) already makes up ~10% of all Prime volume just six months after launch. On Apr 2 2026, Amazon added a 3.5% fuel/logistics surcharge on 3P fulfillment fees (effective Apr 17).
Amazon does not always need to beat Temu on price — it needs to maintain the perception that its ecosystem (fast delivery, trust, selection, returns) justifies a moderate premium. Amazon Haul (launched Nov-2024, now in 25 countries) is more placeholder than disruptor: its strategic value is defensive, preventing Temu and AliExpress from being the only ultra-low-price options.
Against AliExpress, Temu’s Consumer-to-Manufacturer (C2M) model uses algorithmic demand analysis across 90+ countries to drive factory scheduling and real-time dynamic pricing, plus the inherited reverse-auction mechanism — a structural advantage AliExpress’s seller-set-pricing model cannot match.
Shein
Shein’s 1P fast-fashion business is its cash cow — and calling it fast fashion is a disservice; its supplier network (hundreds of factories on an on-demand model) moves from design to finished product in under 7 days. Its US network (5 confirmed facilities, centered on an 800k+ ft² flagship in Whitestown, Indiana) serves most domestic orders in under 7 days. Shein’s commission is a flat 10% referral fee (0% for the first 30 days). The Shein Fulfillment Service mirrors FBA but is restricted to sellers exceeding $5MM in sales. In Sep-2025 it launched Shein Xcelerator, selling its on-demand supply chain as a service (pilot: 22 brands; sellers must keep a Shein storefront).
Walmart
Walmart is a beneficiary of the tariff-driven reshuffling — it already has substantial US infrastructure. Its marketplace surpassed 200,000 sellers in mid-2025, with e-commerce growing 20%+ for four straight quarters (44k sellers added in the first five months of 2025 alone; roughly a third are Chinese). Walmart is absorbing the same Chinese supply base that powers Amazon and Temu, but through a domestically warehoused, duty-paid, faster-delivery model. Its 6%–20% referral with no subscription fee undercuts Amazon.
| Feature | Walmart+ | Amazon Prime (US) |
|---|---|---|
| Annual price | $98 | $139 |
| Monthly price | $12.95 | $14.99 |
| Discounted plans | $49/yr or $6.47/mo (Assist/Student) | $69/yr (Young Adults) / $6.99/mo (Prime Access) |
| Standard shipping | Free next/two-day, no minimum | Free one/two-day, no minimum |
Walmart Fulfillment Services (WFS) now handles 44% of marketplace volume — simpler than FBA, with no monthly storage fees for the first 90 days and no $39.99 professional fee. Walmart calculates referral fees on product price only (Amazon includes shipping). Walmart Connect (US ads) grew 41% in Q4 alone; global ad revenue hit $6.4B in FY26 (+46% YoY; FY25 $4.4B). The dollar gap to Amazon’s $68.6B is still ~11:1, but Walmart’s growth rate is more than double. As CFO John David Rainey put it: “Fully a third of our profit in the most recent quarter was related to advertising and membership income.”
AliExpress
AliExpress is Temu’s most exposed competitor to the regulatory changes — the same tariff regime, but weaker brand recognition, slower delivery, and less aggressive adaptation. In Q4-2025, Alibaba’s international division (AIDC) delivered only +4% YoY revenue, to RMB 37B. The AliExpress Choice program (Cainiao-managed logistics, sub-15-day delivery) drives ~half of orders and runs across 55+ countries, but remains fundamentally cross-border.
Temu (ex-US)
Temu is moving toward an 80% local-to-local goal. As of Dec-2025, the Temu Local Seller Program is live across 37 countries — most of Europe, plus Australia, Brazil, Canada, Japan, Korea, Mexico, Türkiye, the UAE, and the US.
Temu’s global MAU reached roughly 534MM (Dec-25) versus ~379MM (Dec-24), about +41% YoY, with the steepest growth in Latin America and the Gulf. (Full per-country MAU detail — iOS/Android splits across 37 markets — omitted here; the source matrix did not extract reliably and I won’t reproduce numbers I can’t verify cell-by-cell.)
Europe is likely to overtake the US as Temu’s largest market by GMV, if it hasn’t already — 115.7MM EU MAU by Jul-2025. The EU is removing its €150 de minimis, implementing a €3-per-item customs duty effective 1 Jul-2026, with a permanent tariff-from-first-euro regime planned for 2028 (the UK is consulting on removing its £135 threshold). Temu is being proactive, building self-operated warehouses (10 confirmed in Europe as of Oct-2025) across Germany, France, Spain, the Netherlands, Italy, and Austria, with more expected in Poland, Belgium, and Hungary.
TikTok Shop (EU)
TikTok Shop is rationalizing its economics, extracting more from its existing seller base. As of 8 Jan-2026, sellers in Germany, Spain, France, Italy, and Ireland saw fees rise from 5% to 9% per sale — in line with the UK (5% → 9% on 2 Sep-2025). Two partial offsets exist: hitting targets for Live and shoppable videos, or qualifying for a 5% rate (Electronics/Collectibles) after 90 days at 0% with a performance score of 4.0+. In context, 9% still sits at the lower end of European marketplaces (eBay 5%–14.9%; Amazon 7.14%–15.3%).
Case study: Brazil
In South America, Uruguay-based MercadoLibre operates in 18 countries (>50% of revenue from Brazil). Its integrated logistics + fintech + commerce ecosystem presents substantial barriers to entry — the strongest regional moat of any competitor Temu faces. Switching costs are high because customers and sellers rely on the same infrastructure for buying, selling, shipping, and banking.
MercadoLibre operates 30+ fulfillment centers (21 in Brazil by end-2025) plus its own Boeing 737-800BCF freighters. It will invest 57B reais ($10.9B) in Brazil in 2026 (+50% YoY), opening 14 new centers (to 42 total) and adding 10k jobs (to 70k). Its Meli+ subscription has three tiers:
| Country | Tiers (monthly) | Free-delivery threshold | Timeline |
|---|---|---|---|
| Argentina | ARS 3,490 (Esencial) / 7,999 (Total) | ARS 16,000–30,000 | Same-day / 1 day |
| Brazil | BRL 9.90 / 19.9 / 74.9 (Mega) | BRL 19 | Same-day |
| Chile | CLP 2,990 / 7,990 | CLP 9,990–19,990 | 1–2 days |
| Colombia | COP 9,900 / 24,900 | COP 60,000 | 1–2 days |
| Mexico | MXN 59 / 129 | — | 1–3 days |
Like Pinduoduo, ML management is not prioritizing near-term profitability, instead deepening engagement (it monitors DAU growth as the key stickiness indicator). It has cut free-shipping thresholds in 2017, 2020, 2021, and 2025 — each followed by GMV acceleration; the latest (15 Oct-2025) launched free same-day express for Meli+ above R$19 (~$3.90). On 21 Jan-2026 it replaced fixed fees under R$79 with variable charges, and on 31 Mar-2026 it began piloting OTC medication in São Paulo (it bought drugstore chain Target in Aug-2025).
By contrast, Temu entered Brazil in Jun-2024 but faces severe structural constraints: 10+ day average delivery vs ML’s same-day, ~44.5% effective tax (20% import duty + 17% ICMS VAT) on sub-$50 goods, and no integrated credit to rival Mercado Pago.
Amazon (Brazil)
| Country | Status | Monthly | Annual | Free-delivery threshold |
|---|---|---|---|---|
| Brazil | Full Prime | R$19.90 ($4) | R$166.80 ($33.3) | R$149 ($29.8) |
| Colombia | Full Prime | COP 24,900 ($6.9) | COP 165,600 ($46) | — |
| Peru / Chile / Argentina | Prime Video only | ~$5.99 | — | — |
Amazon is expanding aggressively in Brazil: an 89% FBA fee cut for products over R$79 (Aug-2025); zero FBA fees for items over R$100 plus free collection/storage (from Feb 6 2026, extendable through July for sellers allocating 3.5%+ of revenue to Ads); a 15th FC in Santa Maria (Nov-2025, 135k packages/day); and the launch of Amazon Now 15-minute delivery in São Paulo (3 Mar-2026, partnered with Rappi), which reached 8 cities within a week. MercadoLibre’s position is not truly safe.
Shopee
Shopee’s strength is local adaptation — local logistics (SPX Express handles 70%+ of orders) and payment integration in each market. In Brazil (its strongest non-Asian market) it runs 14 distribution centers with 25,000 employees, delivering 25% of Greater São Paulo orders next-day. Sea’s cross-subsidization helps: Garena’s gaming ($2.4B 2025 revenue) and SeaMoney’s fintech ($3.8B revenue, book over $7.8B) provide cash flow, and SeaMoney is being integrated into Shopee for point-of-purchase credit. On 4 Feb-2026 it introduced tiered commissions: 20% + R$4 for items under R$80; 14% + R$16–26 for higher-priced items.
Shein (Brazil)
In Apr-2023, Shein pledged $150MM to build a Brazilian manufacturing/export hub (2,000 local factories, 100k jobs by 2026, targeting 85% local sourcing). The trajectory proved too ambitious — a spokesperson acknowledged it “did not go as expected,” and Reuters found Shein demanded local suppliers cut prices and deliver faster than they could manage. The Guangdong model is not easily replicated. Sellers pay 16% commission plus weight-based logistics fees.
TikTok Shop (Brazil)
TikTok Shop entered Brazil in May-2025; by Dec-2025 it had 131MM local users aged 18+. On 31 Mar-2026, Reuters reported TikTok is seeking Brazilian central-bank approval to operate as a lending and payments fintech; in late 2025 it said it would invest more than 200B reais ($38.4B) in local data centers.
Appendix — margin development
Gross margin % and SG&A as % of sales, quarterly Q1-2016 → Q3-2025, for PDD, BABA, JD, Amazon, eBay, and MercadoLibre.
Sources
- SensorTower — Temu, Pinduoduo, and Brazil e-commerce MAU / DAU and app-download estimates (sensortower.com).
- National Bureau of Statistics of China (NBS) — China online retail sales and e-commerce GMV (stats.gov.cn).
- QuestMobile — Pinduoduo MAU / DAU and DAU/MAU engagement (questmobile.com.cn).
- International Trade Administration — global live-commerce GMV (trade.gov).
- EU Commission — Digital Services Act decisions and penalties (ec.europa.eu/commission/presscorner).
- Reuters — Shein Brazil sourcing; TikTok Brazil fintech filing.
- Company disclosures & earnings calls — PDD (incl. 2024 Form 20-F), Alibaba, JD, Amazon, Walmart, MercadoLibre, Sea, and TikTok Shop / FBT and Amazon FBA published fee schedules.
- Author channel checks and FreightOS benchmark data for cross-border shipping costs.
Moiz Saeed · Himalaya Capital · 1301 2nd Ave, Suite 2100, Seattle, WA 98101