Livestream commerce data in 2026, and which numbers survive a second look
A read of the published livestream shopping figures for 2026, covering the US, China and Southeast Asia, why the market-size forecasts disagree by a factor of a thousand, and the one metric an operator can actually plan against.
Search for the size of the livestream shopping market and you will be handed three numbers for the same year: 2.7 billion dollars, 230 billion, and 2.5 trillion. All three are published, all three are for 2026, and none of them are lying. They are measuring different things and none of them say which.
This is a read of the livestream commerce figures that are actually attributable in 2026: who published them, what they count, and what an operator should do with them.
The United States: real, growing, and smaller than the pitch decks suggest
The most defensible US figures come from EMARKETER, which forecasts rather than resells:
| Metric | Figure | Period |
|---|---|---|
| US livestream commerce sales | $14.64 billion | 2025 |
| Growth on the prior year | close to 50% | 2024 to 2025 |
| Growth in the number of buyers | 21.5% | 2024 to 2025 |
| Share of US digital buyers who bought via livestream | 21.7% | 2025 |
| Livestream share of the US social commerce market | close to 20% | 2025 |
Set against that, EMARKETER's own survey work with Bizrate Insights found 43% of US adults have not used livestream shopping and are not interested in it, and a survey with impact.com put purchases from creator livestreams at 11% of creator-driven shoppers. A category can grow 50% a year and still be a minority behaviour. Both of those are true at once, and a forecast that quotes only the first one is selling something.
The most useful number in the whole dataset is a decline
Ecommerce intelligence firm Charm.io tracks product-level sales on TikTok Shop. Its data, reported by Modern Retail, shows livestream's share of total TikTok Shop sales in the US falling from 26% in July 2024 to 18% in July 2025. Charm's read is that livestreaming has never accounted for more than 30% of revenue on the platform, while pre-recorded creator video has grown into roughly two thirds of it.
That is the opposite of the direction every livestream vendor's homepage implies, and it is the single most important number here. In the US, short video is doing the selling and livestream is doing the closing on a narrower set of categories. Anyone budgeting a US livestream programme on the assumption that the format is about to become the default channel is budgeting against the trend.
China is not a preview of your market, it is a different market
The gap is not subtle. ARK Invest puts livestream at roughly 60% of China's ecommerce market against roughly 5% in the US.
The Chinese figures themselves are strong. A joint study group from the Development Research Center of the State Administration for Market Regulation and the Chinese Academy of Social Sciences reported through Xinhua put 2025 livestream ecommerce GMV above 5 trillion yuan, close to a third of all online retail, with an industry user base expected to reach 660 million. The same study group put livestream's contribution to ecommerce growth in 2024 at 80%, and market size growth between 2019 and 2024 at more than twelvefold. For scale, China's physical-goods online retail was 13.09 trillion yuan in 2025, up 5.2%, per the National Bureau of Statistics.
Read those two paragraphs together and the useful conclusion is not "the West will catch up". It is that livestream share is set by platform design and shopping habit, not by how good the format is. China built discovery, payment and fulfilment around live video. The US bolted live video onto a search and feed habit that already worked.
Southeast Asia is where the growth actually is
Momentum Works' Southeast Asia ecommerce research, summarised by Digital in Asia, puts TikTok Shop's regional GMV at $45.6 billion for 2025, roughly ten times the $4.4 billion of 2022, and around 71% of TikTok Shop's global GMV.
| Market | 2025 GMV | Year on year |
|---|---|---|
| Indonesia | $13.1 billion | +111% |
| Thailand | $10 to 12 billion | +101% |
| Vietnam | $7 to 8 billion | +150% |
| Malaysia | over $5 billion | +132% |
| Philippines | $4 to 5 billion | +99% |
One number in that research deserves more attention than the GMV: average order value on TikTok Shop in the region runs $4.50 to $6.00, against $13 to $15 on Shopee. Live selling in Southeast Asia moves a large number of small baskets. That is an operating model, not a footnote. It decides your margin per order, your shipping economics and, most of all, how many hours you have to be on air.
Why the market-size forecasts disagree by a factor of a thousand
The three numbers at the top of this post are not a scandal, they are three different denominators:
- Platform software revenue. What brands pay vendors for live shopping tooling. Small, in the low billions, and the only one of the three that is really a market in the ordinary sense.
- Livestream GMV. The value of goods sold through live video. Hundreds of billions globally, dominated by China.
- "Live ecommerce" defined broadly. Any ecommerce touched by video anywhere in the funnel. This is how a forecast reaches trillions, and it is close to useless for planning.
The practical rule: never use a market-size figure whose denominator you cannot name in one sentence. If the report will not tell you whether it is counting software fees or goods sold, it is not a measurement.
The same caution applies to the conversion rates that circulate. The widely quoted range of 9% to 30% against 2% to 3% for standard ecommerce originates with a live shopping vendor, Firework. It may well be right for their customers' sessions. It is a vendor benchmark, not an industry measurement, and it is usually reproduced without that label.
The metric that actually plans a business: revenue per stream hour
Everything above is context. The number you can act on is what an hour on air is worth to you, because that is the unit you buy.
Two published data points give you the shape of it. Skincare brand goPure ran 483 hours of TikTok Live for $1 million in revenue and 64 million impressions, which works out around $2,070 per hour. And TikTok said brands and creators hosted more than 8 million hours of livestream shopping in the US in 2024. Set against Pop Mart, where 85% of TikTok Shop sales in June 2025 came from livestreams, the spread is enormous.
That spread is the finding. Livestream performance varies more by catalogue and category than by country. Collectibles, beauty and apparel behave nothing like household goods on the same platform in the same week. Which means the only benchmark worth having is your own, measured on your own catalogue.
What changes when the hour is cheap
The reason so many of these figures plateau is arithmetic. A human-hosted stream costs a host, a studio and a schedule, so it can only be pointed at hours with enough traffic to cover that. Everything either side of peak goes unserved, and so does every language you do not staff.
That is where an AI host changes the shape rather than the rate. Genpio charges by usage, so the relevant question stops being "can we afford a host for this hour" and becomes "does this hour return more than it costs". Streams run overnight and across 19+ platforms, answer chat in 85+ languages in under a second, and run on models we trained and serve ourselves rather than on a chain of rented APIs.
None of that improves a conversion rate on its own. What it does is make the long tail of hours and languages testable, which is the only honest way to find out what your own numbers are.
If you want your own figure rather than someone else's, the shortest route is a paid one-hour trial: load your real catalogue, run a real hour, and measure the one benchmark that transfers.
Sources
- EMARKETER, livestream commerce statistics and FAQ, 2026
- Charm.io data reported by Modern Retail
- Development Research Center of the State Administration for Market Regulation and the Chinese Academy of Social Sciences, via Xinhua, March 2026
- Momentum Works Southeast Asia ecommerce research, summarised by Digital in Asia