A fashion seller chasing 100 million yuan in annual GMV once opened the Douyin ledger and did the ugly math: ads plus commissions ate nearly half of GMV, returns hit about 70%—of 100 million in sales, maybe 30 million remained as “effective” GMV.

That spreadsheet is not an outlier. In 2026, content commerce is no longer a story about posting a video and printing money. Douyin, Kuaishou, and Xiaohongshu are running three different games: the scale leader is defending, the challenger that lost its growth story is attacking, and the reputation marketplace is breaking out. Merchants have one real question—where do the goods and the team go?

Three platforms, three decision scenes

People lump them together as “content commerce,” but shoppers do not.

Platform	Buying mood	What merchants feel
Douyin	Scroll, then buy	Scale + ad/price sensitivity
Kuaishou	Trust + familiar trade	Repurchase stories; weak AOV/brand
Xiaohongshu	Check reputation first	High-ticket decisions, not pure price war

A XinEntropy piece republished by 36Kr puts it bluntly: algorithms cannot copy that mental model. Blasting one SKU and one creative pack across all three usually means winning nowhere. Pick the decision scene first.

Douyin: still the scale arena, now a calculable defense

Douyin remains the volume leader. Public reports put 2025 e-commerce GMV around 4.4 trillion yuan—about 2.75× Kuaishou—and claimed more than 120,000 live sellers doubled GMV year over year during 618. A separate 36Kr exclusive said GMV grew about 34% over a trailing twelve-month window ending mid-2025. Definitions differ; the direction does not: large pie, cooler growth.

XinEntropy’s piece also lays out ad-side pressure:

• Broker estimates in a mid-teens to ~20% band
• Qianchuan CPM: ~80 yuan (2023) → ~120 yuan (2026)
• One beauty brand: 2M yuan monthly ads, ROI ~3× → ~1.5×
• Top hosts’ GMV share ~10%+ — less superstar certainty, more volatility

Platform strategy swung from subsidizing buyers to helping merchants cut cost. 36Kr’s account of platform messaging says the 2026 “nine merchant support” upgrade saved merchants more than 32 billion yuan in 2025; other accounts cite over 6.5 billion yuan in a half-year window on shipping insurance alone and growth for about 300,000 smaller sellers. Ranking favors high settlement and service; risky returns get filtered earlier.

The pitch: stop gambling on traffic; run a forecastable shop.

Infrastructure bets—local-warehouse flagships, short-drama commerce experiments—raise the ticket price from “can shoot video” to “can survive ad inflation, returns, and fulfillment.”

Need explosion and coverage? Douyin is hard to skip. Thin margin and high returns? The main warehouse there ties survival to CPM and return rate.

Kuaishou: growth story faded, shelves and brands as the attack

Public accounts say e-commerce GMV growth once hit about 78% in 2021, then landed near 1.6 trillion yuan in 2025 with growth around 15%. Tencent’s block sale—about HK$12.5 billion cashed out—landed as an emotional signal as much as a financial one.

The counterattack on paper:

1. Reorganize around brand/commercialization, creators, white-label
2. Crack down on scripted diversion and PK gimmicks
3. Push brand penetration and shelf-like commerce
4. Float huge traffic support and “sales hosting” for merchants

New narrative assets can raise giant rounds yet still sit small against core revenue while burning cash. Kuaishou is not abandoning livestream—it is sliding certainty toward shelves, search, hosting, and brand building.

Goods that live on repurchase and community trust still fit. National brand-burst calendars need brandization to land before core elasticity runs out.

Xiaohongshu: sharpest reputation blade, unfinished loop

XinEntropy puts 2025 e-commerce GMV up about 72% year over year; Huxiu cites roughly 850 billion yuan GMV with ads around 32 billion yuan—about 76% of revenue. Numbers disagree; the structure does not: discovery is strong, ads are fat, the commerce loop is still missing the last mile.

Edge	Signal	Caveat
High ticket	>¥3,000 goods said to convert ~47% better than Douyin	High AOV ≠ survival
Viral stories	Amateur livestreams clearing huge GMV in two days	Not the median
Loop gap	On-platform conversion often ~0.7%–1.2%	Decide here, buy elsewhere
Merchant reality	“>60% lose money or barely break even” claims	Seeding ≠ easy profit

Breakout bets—cross-border RedShop, sports rights, AI orgs, IPO talk—raise imagination and bargain against the core discovery habit.

Need taste and high-ticket deliberation? Xiaohongshu is scarce. Need volume and fulfillment? Parking the main warehouse there often overestimates seeding → checkout.

Main warehouse = risk structure

Choice	Common mistake	More realistic split
Brand vs white-label	White-label always wins	Fast cash vs long assets—if you can fund the climb
One vs three	Blast all three	One deep surface—or assign roles, not cloned KPIs
Main warehouse	Put it on the biggest app	Warehouse = inventory, after-sales, scripts, content risk

Pattern emerging: white-label goes deep on one surface; brands assign roles—Douyin for scale, Kuaishou for trust/repurchase, Xiaohongshu for seeding and high-ticket decisions.

• Douyin main warehouse → absorb ad inflation and returns
• Kuaishou main warehouse → shelves/repurchase must succeed livestream
• Xiaohongshu main warehouse → seeding must close on-platform

Content commerce is a pro league now. Winners decide whether a SKU’s decision happens while scrolling, among trusted peers, or inside a reputation search—then place the goods—rather than flooding three seller backends and blaming the algorithm for silence.

Sources

• Three content-commerce platforms in 2026 (36Kr / XinEntropy)
• Douyin GMV +34%, merchant-first strategy (36Kr)
• Xiaohongshu seeding economics (Huxiu)