"Which social platform should we sell on?" is the question every brand asks, and we've answered it directly in our breakdown of which platforms are actually driving sales. But sitting underneath that question is an assumption worth pulling out and examining, because it quietly misdirects a lot of budget: the assumption that the platform is the thing doing the selling.
It mostly isn't. When you break social commerce sales down by where they actually originate, the pattern is consistent across markets and platforms — the storefront is a minority channel, and the majority of revenue moves through individual people talking to their audiences. The platform is the venue. The creator is the engine. Getting that the wrong way round is how brands end up with a beautifully configured shop and almost no sales going through it.
Where the sales actually originate
The clearest evidence comes from TikTok Shop, the platform with the most published data. In analyses of its US market during its first full year of scale, short-form video accounted for well over half of sales, in-app shop browsing a much smaller share, and livestreaming a slice smaller still. Set against that, influencer-made videos and livestreams were credited with the clear majority of US sales — the storefront was never the main event.
More recent third-party estimates put affiliate and creator-driven content at roughly two-fifths of the platform's total sales, and short-form video at close to 60% of them. The figures wobble depending on who's counting, but the direction never does: people sell; storefronts merely hold inventory. A brand that sets up shop and posts from its own account has effectively opened a stall in a busy market and then declined to hire anyone to stand in it.
The reframe Choosing a platform is choosing a venue. It decides who might walk past. It doesn't decide whether anyone sells anything — that's down to who you put on the floor and whether the audience believes them.
The three layers of a social sale
Once you stop treating "the platform" as a single thing, a social commerce sale visibly separates into three jobs. Most brands over-invest in the first and neglect the other two.
| Layer | What it decides | Who controls it | Common mistake |
|---|---|---|---|
| Venue the platform |
Who could see you, and the format | The platform | Treating platform choice as the whole strategy |
| Engine the creator |
Whether anyone is persuaded | The creator, not you | Buying reach instead of credibility |
| Close checkout & fulfilment |
Whether the sale completes, and who owns the customer | You — if you choose well | Defaulting to in-app and renting the relationship |
That third row deserves more attention than it gets. In-app checkout removes friction, which is genuinely valuable, but survey work reported by EMARKETER suggests a meaningful share of younger shoppers still prefer to complete purchases with retailers they already trust to handle payment, delivery and returns. Discovery and checkout don't have to happen in the same place — a pattern we saw repeated with AI shopping assistants, where people were happy to discover in one environment and buy in another. If the close happens on your own site, that's where product page optimisation starts earning its keep.
Katana
Real-time inventory and production management for makers, connecting Shopify, accounting, and your sales channels.
Best for: Scaling makers and ecommerce manufacturers
The trust paradox nobody prices in
Here's the tension at the centre of creator-led commerce, and it rarely appears in the enthusiastic write-ups. The engine driving these sales is trusted less than the advertising it replaced.
Research from the National Advertising Division, cited by EMARKETER, found that around a quarter of consumers say they don't trust influencer marketing — roughly double the share who say the same about advertising generally. The same body of work found a large majority distrust creators who don't disclose their brand relationships. Yet well over half of adults report having bought something because a creator recommended it. Both things are true at once: creators are the most effective sales mechanism on social platforms and the least trusted one, and the gap between those facts is where brand damage happens.
What makes that riskier is how little diligence sits behind the average partnership. Research from EMARKETER and Viral Nation found more than half of marketers spend half an hour or less vetting an individual creator. Half an hour, to select the person who will be the entire persuasive engine of a campaign — and whose past posts, undisclosed deals or reputation become yours by association the moment you hand them a discount code.
→ Disclosure history. Do they label paid posts consistently? Non-disclosure is the single biggest trust-killer with audiences.
→ Audience overlap, not audience size. Followers are not buyers. Who actually engages, and are they your customer?
→ Category conflicts. Have they promoted a direct competitor in the last quarter, or three?
→ Comment sentiment. Read the replies on their last ten posts. Audiences say plainly whether they believe someone.
→ Sales history, not reach history. Ask for conversion data on comparable products, not a media kit.
An hour of this per creator is cheap. Discovering it afterwards is not.
The uncomfortable corollary is that the credibility you're renting was built by someone else and can be spent by them without asking you. This is the same fragility we've written about with everyday creators outperforming celebrity endorsements: the trust is the asset, and it is not yours.
Live commerce: best conversion, hardest operations
Livestream selling is the format that most clearly demonstrates the creator-as-engine thesis, because there is no storefront involved at all — just a person, in real time, selling. Industry estimates put its conversion rate at roughly triple that of ordinary feed content, and it has grown from a small fraction of TikTok Shop's sales into a substantial share of them within about two years.
Two honest caveats. First, live is operationally brutal: consistent sellers stream many hours a week, and the economics only work when you can sustain that frequency with products that genuinely demonstrate well. It is a staffing commitment, not a campaign. Second, and more encouragingly, returns from live purchases appear meaningfully lower than standard e-commerce returns — which makes intuitive sense, since someone who watched a product used for twenty minutes has fewer surprises coming than someone who bought from a single photograph. If you're weighing live against other video formats, our piece on shoppable video on CTV covers the same trade-off in a different setting.
The concentration nobody advertises
One more thing worth knowing before you build a plan on this channel. The "anyone can sell on social" story is, in revenue terms, misleading. Analysis from Marketplace Pulse has found that the top 1% of US TikTok Shop sellers take roughly 60% of all sales, with the top fraction of a percent taking over a quarter on their own.
That's a winner-takes-most market wearing the costume of a democratic one. It doesn't mean don't participate — it means don't budget as though median outcomes are typical, because they aren't. The realistic goal for most brands isn't to become a top-1% seller; it's to work with the creators who already are, or to win a narrow category where concentration hasn't set in yet.
A note on the numbers
Read three articles about the size of this market and you'll get three incompatible figures. Forecasts for TikTok Shop alone in 2026 range from roughly $23bn to $27bn for US sales, and from $53bn to over $110bn globally, depending on the source. Some of that spread is genuine uncertainty; much of it is definitional — US versus global, gross merchandise value versus net e-commerce sales, forecast versus completed year — and a good deal is SEO-driven stat pages copying each other without checking.
The practical response is the one we apply to any headline multiple: use the numbers to establish direction, not to build a forecast. The mechanism here is well evidenced and stable — creators convert, storefronts hold stock, live converts best and costs most. The precise size of the pie changes every time someone re-measures it.
If you'd rather have the creator programme, the measurement and the checkout path designed as one system rather than three disconnected experiments, that's what e-commerce marketing support is for.
The honest summary
Social commerce is real and large, but it isn't the platforms doing the selling — it's the people on them, and the sooner a brand internalises that, the better its budget performs. Choose a venue by all means, then spend the real effort on the engine: which creators, chosen for credibility rather than reach, vetted properly rather than in half an hour. Decide deliberately where the purchase completes instead of defaulting to in-app and handing the customer relationship to a platform. Treat live as the high-conversion, high-labour format it genuinely is rather than a quick win. And hold the market-size figures loosely, because they contradict each other freely. The durable insight underneath all of it is unglamorous: this channel runs on borrowed trust, and every decision worth making is really a decision about who you're borrowing it from.
Selling on social but not seeing the sales?
We build creator programmes, measurement and checkout paths that work as one system.
Explore E-commerce Marketing →Frequently asked questions
Do creators really drive social commerce sales?
The available breakdowns suggest they drive most of them. On TikTok Shop, analyses of the US market have credited influencer videos and livestreams with the clear majority of sales, and short-form video consistently accounts for a far larger share than in-app shop browsing. The practical reading: the platform is a venue, creators are the sales engine — so a storefront without a creator programme is a shop with nobody selling in it.
Which platform drives the most social commerce sales in 2026?
By growth and e-commerce sales, TikTok Shop leads in the US, with EMARKETER forecasting it will run a larger US e-commerce business than several major national retailers this year. Facebook still holds the largest base of social buyers. But platform choice matters less than most guides imply, because within any platform sales concentrate around creator-led content rather than spreading evenly across brand accounts.
Is livestream shopping worth doing?
It converts better than any other social format and costs far more to run. Industry estimates put live conversion at roughly triple feed content, and live has grown into a substantial share of TikTok Shop's sales. The catch is operational: consistent sellers stream many hours weekly, so it pays only with sustained frequency and products that demonstrate well. Returns tend to be lower, since buyers who watch a demonstration get fewer surprises.
Should you sell in-app or send shoppers to your own site?
Use social for discovery and think carefully about where the purchase completes. In-app checkout removes friction, but a meaningful share of younger shoppers still prefer buying through retailers they already trust for payment, shipping and returns. In-app selling also hands the customer relationship to the platform. Many brands do best letting creators drive discovery while keeping the close, and the data, on channels they control.
