Six markets, six stages
Creator commerce across Southeast Asia is not one channel running at different speeds in different countries. Each of the six markets covered here has reached a distinct structural stage. A brief built for one will consistently underperform when applied to another without adjustment.
Thailand is the most structurally mature. TikTok Shop’s Thailand GMV reached roughly USD 5.9 billion in the first half of 2025, one of the largest single-country platform GMV figures in the region for that period. Thailand’s top creators by revenue skew heavily toward Key Opinion Sellers, creators who convert directly inside a livestream session, rather than Key Opinion Leaders who build awareness without a direct sales function. A standard KOL-first awareness brief run in Thailand is one stage behind where the market has already moved.
Vietnam follows close behind on infrastructure depth, though for a different reason. Two platforms, Shopee and TikTok Shop, control 97% of the country’s ecommerce GMV. That concentration makes the creator strategy more legible than in any other SEA market covered here: a brand knows exactly which two platform relationships matter, with no ambiguity about where to concentrate creator budget.
Indonesia carries the region’s largest raw volume, with video and livestream formats accounting for a substantial and growing share of online GMV. The market’s infrastructure remains concentrated around Jakarta, which means national penetration figures overstate how evenly that volume is actually distributed. A creator brief built for Jakarta buyers does not automatically translate to buyers in other Indonesian cities.
Malaysia is recovering and accelerating. Livestream commerce has become a meaningful share of platform GMV following TikTok Shop’s continued market growth, Malaysian spend on the platform reportedly reaching RM12.07 billion in 2024, up 104% year on year. Mid-tier category specialist creators are reportedly outperforming general-audience creators in this environment.
The Philippines runs high creator penetration on a comparatively thin logistics and infrastructure base. Short video content converts more reliably than livestream formats at the market’s current maturity level, a pattern that echoes Singapore’s stage more than it echoes Thailand’s or Indonesia’s.
Singapore sits last on live commerce adoption among the six, with a smaller, more premium-skewed creator base. It functions more usefully as a content production base for regional deployment, hero content shot once and adapted across markets, than as a standalone live commerce market in its own right.
Where AI is changing the economics
Brief inconsistency, not creator quality, is reportedly the largest driver of variable creator performance under otherwise identical commission structures across these six markets. AI-assisted brief generation is starting to reduce that variability by keeping brief quality consistent across a larger roster of creators.
Content adaptation is a second area of change. Hero content produced in Thai or Vietnamese can be adapted into Bahasa or Tagalog for regional deployment at a fraction of the cost of separate market-by-market production. AI-driven hosting is also starting to cover off-peak catalogue hours in livestream formats without a proportional increase in production cost. Performance pattern recognition is shifting from a post-campaign review exercise to a mid-campaign adjustment tool in more advanced programmes.
Thailand and Vietnam are adopting these tools fastest, consistent with their more mature underlying infrastructure. The gap between markets has little to do with budget. It comes down to the distance between what each platform’s tooling already supports and what most brand operating models are actually built to use.
The conditional recommendation
Match the creator brief format to the market’s actual stage, not to the format that worked in the brand’s home market or largest existing market. A KOS-first, livestream-heavy brief is the right choice for Thailand and increasingly Malaysia. A KOL-first, short video brief is the right choice for the Philippines and Singapore. Vietnam and Indonesia sit in between, where the platform relationship itself, not the creator format, is the primary strategic decision. Building one regional brief and running it unchanged across all six markets is the single most common execution error in this channel.
Research Ledger (5 claims)
| Entity | Claim | Source | Status |
|---|---|---|---|
| Thailand | TikTok Shop GMV reached approximately USD 5.9 billion in H1 2025, among the largest single-platform GMV figures in the region for that period | Multiple GMV tracker summaries citing TikTok Shop Thailand performance (marketmaze.me, hiregrowth.ai) | VERIFIED |
| Thailand | Most top creators by platform revenue operate as Key Opinion Sellers rather than Key Opinion Leaders | Referenced across creator commerce maturity analyses, consistent with Thailand’s mature live commerce infrastructure | INFERRED (directionally well supported by Thailand’s live commerce maturity, the specific “nine of the top ten” ratio was not independently reproduced from a named primary leaderboard in this pass) |
| Vietnam | Two platforms, Shopee and TikTok Shop, control 97% of ecommerce GMV | Metric data via theinvestor.vn, kr-asia.com | VERIFIED |
| Indonesia | Video commerce and livestream purchasing represent a substantial share of online GMV, with infrastructure concentrated around Jakarta | Momentum Works and Magpie IQ Indonesia ecommerce analyses | INFERRED (directional concentration and Jakarta-centric infrastructure well supported, the specific 20% video commerce and 60% livestream-purchase figures were not independently reproduced from a single primary table in this pass) |
| Malaysia | Livestream formats generate a meaningful share of platform GMV following TikTok Shop’s continued growth in the market | worldofbuzz.com reporting on Malaysian TikTok Shop spend, RM12.07 billion in 2024, up 104% year on year | INFERRED (growth and livestream relevance confirmed, the specific 22% GMV share figure was not independently reproduced from a primary source in this pass) |