The influencer economy and live commerce: who they are, where they cluster, how they earn, and whether borders exist
Influence flows without borders.Algorithms distribute content by interest, not geography. A creator in New York reaches a consumer in San Francisco effortlessly.
Goods flow with friction.The end of de minimis, the EU customs duty and platform rules made cross-border commerce markedly more expensive in 2025–26.
Live selling barely flows at all.Real-time language, culture and separate platforms (Douyin ≠ TikTok) make livestream selling the most local of all digital formats.
Executive summary
Scale. Goldman Sachs sized the global creator economy at $250 billion in 2023 and sees it approaching ~$480 billion by 2027. China's live-commerce market was worth roughly $900 billion in 2025, close to the size of all US e-commerce. TikTok Shop US sold $15.82 billion in 2025 (eMarketer), with ~$23.4 billion projected for 2026.
Money. In China the "pit fee (坑位费) + commission" model dominates; top hosts generated billions in sales, but the state brought the sector to heel with large tax penalties in 2021–2024 (1.34 billion yuan for Viya). In the West income is extremely unequal: more than half of creators earn under $15,000 a year.
Borders. A Shanghai-based live-commerce producer influences the US consumer not directly but through three indirect channels: as a supplier (feeding US affiliate creators), through format arbitrage (formats proven in China reach the West 12–24 months later), and as an operator (Chinese MCNs incorporating in the US and running local creator networks).
What is happening in Asia is not a threat to Western retail; it is a preview.Emilie Darolles, President Western Europe, NielsenIQ — The Commerce Revolution: Where East Meets West, 2026
The scale of this world, on one logarithmic axis
Billions of US dollars. Different definitions (TAM, GMV, ad spend) should not be conflated; the chart only shows orders of magnitude.
1. Who are the actors?
Core roles
There is no single "influencer" job. Different roles run different business models:
- Content creators are tiered by followers: nano (1K–10K), micro (10K–100K), macro (100K–1M), mega (1M+). Most brand budgets go to the micro tier; on Instagram micros deliver roughly 3x the engagement of megas.
- Live-commerce hosts (主播, zhǔbò) are professionals who sell products in real time. Their Western counterparts are TikTok Shop LIVE and Whatnot hosts.
- Affiliate creators earn a commission per sale. They drive 42% of US TikTok Shop GMV; the top 0.5% of creators (~4,000 accounts) generate 38% of all affiliate revenue.
- MCNs (Multi-Channel Networks) are Chinese agencies that manage, train and supply creators at scale, with studios and supply chains.
- TP operators (Tmall Partners) run brands' Chinese storefronts end to end.
How a Chinese MCN differs from a Western agency
A Western agency mainly represents and negotiates. A Chinese MCN is a vertically integrated factory: talent scouting, training, studios, product pools (货盘), traffic buying and after-sales logistics under one roof. Wuyou Media manages about 90,000 contracted KOLs. Chinese platforms now display the MCN's name on KOL pages, revealing that many "independent" creators belong to large agencies.
| Organisation | Base | Notable |
|---|---|---|
| MeiONE (美ONE) | Shanghai | Li Jiaqi's company; #1 in Hurun 2025 ranking; 1,600+ brands, 250 million viewers in one stream |
| Wuyou Media (无忧传媒) | Beijing/Hangzhou | The "influencer factory"; ~90,000 KOLs, 1.8 billion total followers |
| Yaowang / Yowant (遥望科技) | Hangzhou | Publicly listed; 100+ hosts, 30+ celebrities; 2024 GMV ¥4.5 billion |
| Qianxun (谦寻) | Hangzhou | Taobao Live's largest agency, which produced Viya; 20,000+ brands |
| Three Sheep (三只羊) | Hefei | Founded by Xiao Yangge; rose fast and was fined in 2024 |
| Whatnot | Los Angeles | Collectibles live auctions; 2025 GMV $8 billion, $20 billion valuation in August 2026 |
| TalkShopLive | Los Angeles | Celebrity/brand embeddable video commerce; Amazon Live and Walmart partnerships |
2. Geographic concentration: where and why?
China: Hangzhou is the capital of live commerce
According to the Zhejiang commerce department, Hangzhou hosts ~50,000 hosts and more than 5,000 registered livestream companies; the sector employs over a million people directly or indirectly — one in 244 residents is a host, one in 12 works in a related industry. The landmark is Lijing International Center, the "Influencer Tower". Why Hangzhou? It is the home of Alibaba and Taobao Live, has the densest MCN cluster, and offers new e-commerce firms subsidies of up to 5 million yuan. Suzhou pays hosts who sell over 50 million yuan a year and pay tax locally up to 1 million yuan in cash.
Other hubs: Guangzhou and Shenzhen (supply chain, apparel and electronics), Shanghai (international brand HQs), Beijing (ByteDance/Kuaishou). Clustering follows supply-chain proximity, studio infrastructure, platform headquarters and incentives.
United States: Los Angeles and New York, then Miami and Atlanta
In Ubiquitous Influence's sample of the 200 most prominent US influencers, California leads with 31%. Influencity finds 64% of mega-influencers in California and 20% in New York. The important nuance: this is a concentration of production, not of influence. Content is shot in LA and bought in Ohio.
Where do the 200 most visible US influencers live?
Share of the sample. Small sample; the ordering is reliable, the decimals are not.
Other markets
Southeast Asia is TikTok Shop's growth engine: $45.6 billion GMV in 2025; Indonesia at $13.1 billion is the second-largest market after the US (Momentum Works). Asia-Pacific accounts for 66.8% of global live-commerce revenue. South Korea is mature; Vietnam and Thailand are growing fast; Brazil and India are emerging. The UK is Whatnot's strongest market outside the US. No verifiable GMV data was found for Turkey, so the report leaves it out with caution.
3. How do they make money?
Creator revenue streams in the West
According to Goldman Sachs, brand deals make up ~70% of creator income. A finer breakdown:
Sources of creator income (2026)
Share of total creator revenue.
| Revenue stream | Typical rate / mechanism |
|---|---|
| Brand sponsorship | Micro: $100–1,000 per post; macro: $5,000+ |
| Affiliate commission | TikTok Shop US average 13.02%; open collaboration 10–15%, targeted collaboration 15–50% |
| Platform payouts | YouTube Partner Program, TikTok Creator Rewards, Instagram bonuses |
| Live gifting | Virtual gifts; platform take up to 69% on some platforms |
| Subscriptions | Substack (5 million+ paid subscriptions, 2025), Patreon |
| Own brand / digital products | 45% of full-time creators have launched their own brand |
The income reality: a power law
Share of creators by annual income band.
Live-commerce economics in China
Two components: the pit fee / slot fee (坑位费), a fixed upfront fee the brand pays to place a product in a stream (80,000–100,000 yuan at MCNs; hundreds of thousands for top hosts), and commission (10–30% of sales; 40–50% for some top hosts). Usually both, plus a demand for a deep discount. In 2024–25 the industry shifted toward performance: Taobao Live now collects the slot fee only if sales beat target by 20%.
We sold a lot, but we weren't always profitable, because we sold too cheaply.Damian Maib, founder of Genuine German, on a Douyin campaign that cost 300,000 yuan fixed + 25% commission + a 40% discount
Top hosts and the 2021–2026 crackdowns
- December 2021
Viya (Huang Wei) fined a record 1.34 billion yuan ($210 million) for tax evasion; her accounts were deleted. The same year Cherie was fined 65.55 million and Lin Shanshan 27.67 million yuan. - November 2021
Li Jiaqi ("Lipstick King") sold 10.6 billion yuan+ on Singles' Day; ~19 billion yuan combined with Viya on day one. - 2023–2024
Dong Yuhui, a former English teacher, topped the 2024 net-income ranking at 2.8 billion yuan; he left East Buy and bought out his own channel. - September 2024
Xiao Yangge's Three Sheep group fined 68.95 million yuan over a fake "Hong Kong made" mooncake; accounts of ~2,000 hosts suspended. - 1 September 2026
New regulations governing MCNs took effect.
The big picture: state pressure ended the "super-host" era; brands and platforms learned the risk of depending on a single star.
The seller side: high GMV, thin margins
- In the US the top 1% of sellers drive ~60% of GMV; more than half of the 800,000+ registered shops have never recorded a sale.
- Nominal commission misleads: a 15% commission plus platform fees, shipping and returns can cost a seller ~26.6% of net revenue; brands keep roughly 67% of reported GMV.
- TikTok's seller commission rose from 2% to 4% on 1 January 2026 (~6% in some categories); cross-border sellers in the US must post a $1,500 security deposit per shop.
4. The border question: does geography constrain influence?
There is no single border. There are four, and they are of very different heights. That is the source of the confusion.
The four-border matrix
| Layer | Permeability | Evidence |
|---|---|---|
| a) Attention / algorithm | High | Recommendation engines distribute by interest and engagement. At national scale geography barely matters. Exceptions: inherently local categories such as restaurants, real estate and local services. |
| b) Language / culture | Low | A recorded video crosses with subtitles; a livestream does not. Improvisation, jokes, haggling and replying to comments need fluent language and cultural code. Live commerce is the most local of formats. |
| c) Platform | Medium / low | Douyin ≠ TikTok: separate apps, content and data. TikTok Shop runs country by country; the US requires a legal entity + EIN + US bank + local warehouse. The TikTok US joint venture (January 2026) is separating data from China. |
| d) Trade / regulation | Closing | US de minimis eliminated (May/August 2025); EU €3 customs duty (July 2026) plus proposed €2 fee; UK and Southeast Asian tax changes. |
a) The attention and algorithm layer
Modern recommendation engines distribute content by interest and engagement signals. A New York creator's video reaches a San Francisco consumer; the question itself dissolves, because that creator's audience is a niche, not a city. Only in hyper-local categories (restaurants, real estate, hair salons) are small local accounts worth more than large ones.
b) The language and culture layer
Because it needs real-time conversation, cultural references and fluency, live commerce is the most locally bound of all digital formats. Top Chinese hosts expanding abroad must build local-language teams. English content has the widest cross-language reach; other languages mostly stay within their own regions.
c) The platform layer
- Douyin ≠ TikTok. Separate apps from the same company; a Douyin host cannot stream directly on TikTok.
- TikTok Shop is country by country. Sites practically accessible to Chinese sellers in 2026: six Southeast Asian countries, the UK, the US, Spain, Mexico and a few invitation-only test markets (探行, 2026).
- US entry barrier. A cross-border seller needs a US-registered entity, a US EIN, address and bank account, and a US-based warehouse because domestic returns are mandatory within 30 days. Sole-proprietor licences are increasingly rejected in the US/UK in 2026 (探行).
- The TikTok US joint venture. Under the executive order of 25 September 2025, ByteDance sold more than 80% of its US assets; "TikTok USDS Joint Venture LLC" was formed on 22 January 2026. Oracle, Silver Lake and MGX hold 15% each; ByteDance 19.9%. US user data sits in Oracle's US cloud and the algorithm is being retrained on US data. Meaning: US commerce and data are being legally separated from the Chinese parent.
- Alternative channels. Temu, Shein and AliExpress ran on direct China-to-consumer parcels; that model took a hit in 2025.
d) The trade and regulation layer
- 2 May 2025 — US
The $800 de minimis exemption ended for goods from China and Hong Kong (EO 14256). Postal shipments face a flat option of $100 per item, $200 from June. - 29 August 2025 — US
De minimis suspended for all countries (EO 14324); extended in February 2026. Previously ~4 million parcels a day entered duty-free; 76% of them came from China in FY2024. - 13 December 2024 — EU
GPSR in force: non-EU sellers cannot sell without an EU-based "Responsible Person"; marketplaces remove non-compliant listings. - 1 July 2026 — EU (final)
A temporary €3 customs duty per item type on consignments under €150 (Council Regulation (EU) 2026/382), valid until 2028. 4.6 billion low-value parcels entered the EU in 2024, 91% from China. - ~November 2026 — EU (proposed)
A €2 handling fee per parcel; ECOFIN agreed in principle, final legislation pending. - November 2025 — UK
The Autumn Budget announced removal of the customs duty relief on imports under £135 (~2028–29). VAT has applied to every import since 2021. - 2025–2026 — Southeast Asia
Vietnam applies 10% VAT to low-value imports from 18 February 2025; Thailand applies duty + 7% VAT on all imports from 1 January 2026; Indonesia pushes cross-border sellers toward local licences.
The hit to Temu and Shein in the US (March → May/June 2025)
Percentage change, immediately after de minimis ended for China.
Direct answer: does a Shanghai live-commerce producer influence the US consumer?
Not much directly; but strongly through three indirect channels.
- As a supplier. Chinese manufacturers feed US affiliate creators with products. 42% of US TikTok Shop GMV is affiliate-driven, and most of those products come from Chinese supply chains. Beauty and personal care (TikTok Shop US's largest vertical, ~22%) is the typical case.
- Format arbitrage. Formats proven in China reach the West 12–24 months later: live auctions and countdown bidding (TikTok Shop expanded Countdown Bidding in May 2026), flash sales, store-run streams (店播), automated ads such as GMV Max.
- As an operator. Chinese MCNs and operations firms incorporate in the US and run local creator networks. The visible face is American; the operation behind it is in Shenzhen. Temu's "local-to-local" pivot is the same logic.
5. Market size and data quality
| Metric | Value | Source / year |
|---|---|---|
| Creator economy TAM | $250bn → ~$480bn | Goldman Sachs Research, 2023 → 2027 projection |
| Global influencer marketing spend | $32.55bn → ~$40.5bn | Statista / Influencer Marketing Hub, 2025 → 2026 |
| China live commerce | ~$900bn | NielsenIQ, 2025 (cited market estimate) |
| China live commerce (alternative) | ~$1.1tn | Statista and others, 2026 projection |
| Global live commerce | $172.9bn | Grand View Research, 2025 |
| TikTok Shop US | $15.82bn → ~$23.4bn | eMarketer, 2025 → 2026 projection |
| TikTok Shop global GMV | $64.3bn → ~$112bn | Momentum Works, February 2026 |
TikTok Shop US: sales and the share of live
Bars are GMV in $ billions (left axis); the line is live's share of GMV (right axis). 2026 is a projection.
The East–West gap: live's share of e-commerce
Percent. In China live is the main channel; in the US it is still a small slice.
Consumers who have never bought via social media
Percent, NIQ Consumer Outlook survey.
6. Consumer-side mechanics
The mechanism is not advertising but parasocial trust: the viewer forms a one-way yet emotionally real relationship with someone they have never met, which moves a recommendation from the "ad" category to the "friend's advice" category. Live adds social proof (thousands of concurrent viewers, "500 sold in 20 minutes") and urgency/scarcity (countdowns, "this price only in this stream"); awareness, evaluation, trust and purchase collapse into one session.
Conversion rates: why live is so powerful
Percent, TikTok platform and general e-commerce comparison (2026).
Whom does Gen Z trust?
Share of Gen Z trusting a product recommendation.
Categories and the "sinking market"
Beauty and personal care is TikTok Shop US's largest vertical; fashion leads in shop count; collectibles and trading cards are Whatnot's core. In China, Kuaishou serves lower-tier cities and older users (the "sinking market", 下沉市场); second- and third-tier cities are where live-commerce audiences are densest.
7. 2025–2026 inflection points
AI digital hosts
On JD.com more than 70,000 merchants use AI digital-human hosts; cost has fallen to a tenth of a human host. On 15 June 2025 Luo Yonghao's Baidu-powered avatar sold 55 million yuan ($7.65 million) in 6+ hours — more than his own live stream. Advantages: 24-hour streaming, no scandal risk, multilingual. Disadvantage: it cannot yet carry Li Jiaqi-style "hype" energy.
Platform automation and regulation
- Automated ad tools such as GMV Max are becoming effectively mandatory on TikTok Shop.
- Tax and false-advertising penalties continue in China; MCN regulation took effect on 1 September 2026.
- In the US the FTC enforces sponsorship disclosure; the TikTok US joint venture closed in January 2026.
- In the West, Meta and YouTube restructured live shopping; Whatnot found product-market fit in collectibles and grew ($20 billion valuation in August 2026).
The next frontier: AI shopping agents
Alipay AI Pay processed 120 million transactions in a single week (February 2026). What happens to the influencer economy when an agent, not a human, makes the decision is an open question.
8. How to observe this yourself
- Open TikTok Shop LIVE. Join a beauty or home-goods stream; watch the countdown, the pinned product and the chat dynamics.
- Download Whatnot. Join a trading-card auction; understand why people spend 80–95 minutes a day.
- Compare Douyin. Search "Li Jiaqi livestream" on YouTube; see the difference in scale and professionalism.
- Follow the affiliate economy. Look at LTK, ShopMy or a creator's "shop" links.
- Track format arbitrage. Note a new format in China; check whether it appears in the West 12–24 months later.
- Data literacy. For every statistic ask: TAM, GMV or real revenue? Primary source or SEO blog?
- Follow customs news. Changes like de minimis and the EU €3 duty decide directly who survives.
Recommendations and decision thresholds
- Learning (0–1 month). 2–3 hours a week of active observation on TikTok Shop LIVE and Whatnot; pick one beauty and one collectibles category.
- If you are a seller or brand (1–3 months). Run an open-collaboration test with 5–10 micro creators at 15–20% commission. Threshold: positive ROAS within 30 days → move to targeted collaboration. Compute net margin after the full fee stack (~67% retained), not nominal commission.
- If considering cross-border. Do not start in the US without an LLC + EIN + 3PL; with de minimis gone, direct shipping has no margin. For the EU, watch whether the €2 fee is finalised before November 2026.
- Format tracking (ongoing). China's AI-host and automated-ad trends will reach your market within 12–24 months.
Thresholds that should change a decision: if live conversion drops below 7%, revisit format/host/product fit; if the EU €2 fee is finalised, move to a local-warehouse model; if AI-host cost reaches 10% of a human host, evaluate automation for repetitive stream segments.
Caveats and limitations
- Creator-economy estimates range from $205bn to $480bn and China live commerce from $900bn to $1.1tn; definitional differences are large.
- 2026 GMV figures, the EU €2 fee and the UK customs change are forward-looking projections or proposals, not settled facts.
- Chinese government and platform data (Hangzhou employment, AI-host GMV) have not been independently audited.
- Temu/Shein user data (Sensor Tower, FT) is solid; GMV percentage declines are analyst commentary.
- Some charts use SEO-oriented statistics sites that relay primary sources; read those figures as approximate.
Sources
Primary / strong
- NielsenIQ — China's ~$900B Live-Commerce Market Now Approaches US E-Commerce Scale (July 2026)
- NielsenIQ — 120 Million AI Pay Transactions in a Single Week (2026)
- Goldman Sachs Research — The creator economy could approach half-a-trillion dollars by 2027 (relayed)
- Global Times — China's top livestreamer fined $210m for dodging tax (December 2021)
- South China Morning Post — Hangzhou to provide subsidies for e-commerce industry
- Hangzhou Municipal Government — Hangzhou puts livestreaming e-commerce regulations on the table (2023)
- China Media Project — Hangzhou's Tower of Influence (November 2024)
- Newsweek — Map Shows Which States Have the Most Influencers (July 2024)
- Influencity — The Largest Influencer Study in the US 2023
- NBC / CNBC — AI avatars in China just proved they are ace influencers (June 2025)
- China Daily — Digital humans show future of livestreaming e-commerce (July 2025)
- Grand View Research — Live Commerce Market Size & Share Report
- Dao Insights — Chinese social media start displaying MCNs' names
- Wikipedia — Qianxun Group
Trade and regulation
- ExFreight — De Minimis Rule: End of the $800 Threshold for China
- HLC — De minimis exception to end in August 2025
- 3PL Center — TikTok Shop Tariffs: The Impact on E-Commerce and Fast Fashion (April 2026)
- Tianxia Gongchang Research — China Cross-Border E-Commerce 2026
- Ownswho — Who Owns TikTok? Oracle, Silver Lake, MGX and ByteDance (2026)
- Social Tale — TikTok Shop Cross-Border Selling Guide
- Canopy Management — TikTok Shop Eligibility 2026
- 探行 — Which TikTok Shop Sites Can Chinese Sellers Join in 2026? and Cross-Border Requirements: The Hidden Risks
Industry compilations (use with care)
- Hamstergarage — TikTok Shop Affiliate Statistics 2026 and Affiliate Commission Rates
- TTS Vibes — TikTok Live Shopping Conversion Rate 2026
- Colaba — TikTok Shop LIVE 2026
- Videowise — China Live Commerce Market Size
- CO Consulting — 42 Creator Economy Statistics for 2026
- Self Employed — A Creator Middle Class Is Emerging
- SaaSUltra — The Creator Economy in 2026
- Revenue Memo — Gen Z marketing statistics for 2026
- The World Data — TikTok Shop Statistics 2026
- Branvas — TikTok Shop Statistics 2026
- Stellagent — JD.com's AI Digital Human Sellers Surpass 70,000
- ChoZan — A List of 14 Major MCNs in China
- China Marketing Corp — China Live Streaming E-commerce: Part 1