Temu and Shein: You Are Being Asked to Price Against an Arbitrage That No Longer Exists
Not another rise-of-Temu explainer. The listing decision, taken seriously: what happens to your price architecture if you join, what happens to your discovery if you do not, and why the brand-safety debate is the least useful lens available.
Annie Chan··14 min read
Every board that has had this conversation has had roughly the same version of it. Someone from commercial says the traffic is enormous and free. Someone from brand says it will cheapen us. The CFO asks what the margin looks like. Nobody has good numbers, so the decision gets deferred to next quarter, and it has now been deferred for about nine quarters running.
The debate is stuck because it is being conducted in the wrong terms. Brand safety is a real consideration and it is a recoverable one: a brand that lists somewhere unflattering can stop, and the damage decays. Price architecture is not recoverable in the same way. Once a reference price for your product exists in the market, it exists, and it will be surfaced to your customers by systems you do not control.
The question is not whether Temu or Shein is beneath you. It is whether you can afford to publish a price there that will then be used to evaluate every other price you publish.
The thing that changed, which most brand teams have not registered
Both platforms grew on a structure that has been dismantled. Goods were made in China, sold direct to a Western consumer, and shipped as an individual low-value parcel that entered the destination country without formal customs entry and, in most cases, without duty. That was not a logistics innovation. It was a regulatory position, and it was worth more than any of the operational cleverness layered on top of it.
The United States suspended its de minimis exemption for China and Hong Kong in May 2025, extended the suspension to all countries in August 2025, and in June 2026 US Customs and Border Protection made the suspension indefinite by regulation. Congress has repealed the underlying statutory privilege for commercial shipments with effect from July 2027. The European Union is removing its own EUR 150 duty exemption on a parallel track. The mechanics are covered in detail here.
Both platforms responded the same way: by moving inventory into the destination market and recruiting local sellers. Temu pushed hard into a local fulfilment model, sometimes described as semi-managed, where the seller holds stock inside the destination country, handles domestic fulfilment, and takes more control of pricing, while the platform supplies the storefront, the traffic and the checkout.
That is a sensible response. It is also a confession. A marketplace whose defining advantage was shipping from China has stopped shipping from China.
What the numbers say about the transition
Shein's Hong Kong listing process has forced disclosure that was previously unavailable, and the picture is of a business that grew through the arbitrage and has decelerated sharply without it. Reported revenue reached $41.8 billion in 2025, up 8 percent. In the first quarter of 2026, reported revenue growth slowed to 1.1 percent and the company posted a net loss of $99 million.
The valuation trajectory tells the same story more bluntly. Shein has been reported as seeking a $30 billion to $40 billion valuation in Hong Kong, against a $98.2 billion valuation at an earlier private round.
Temu's parent does not break the business out in a way that supports the same reading, but the regulatory ledger is public. In May 2026 the European Commission fined Temu EUR 200 million under the Digital Services Act for failing to diligently identify, analyse and assess the risk of illegal products reaching EU consumers, with the Commission citing safety failures in tested products including baby toys. It was the largest DSA fine issued to that point. Temu was given until late August 2026 to submit a remediation plan.
Shein's European position has been worse. French authorities moved to suspend the platform in November 2025 following the discovery of unlawful listings, and the company has accumulated French penalties reported to exceed EUR 210 million, including a further EUR 22 million penalty in June 2026 covering consumer-rights, order-confirmation and disclosure failures. A French appeals court declined to uphold a ban in early 2026, which counts as a win, and the partnership with the BHV Marais department store ended after other brands withdrew in protest at Shein's presence.
"Growth of 1.1 percent, a quarterly loss, a valuation cut of roughly two thirds, and nine figures of European fines. Whatever you think you are joining, you are not joining the company from the 2023 case studies."
The trap: you are quoted against a price that is no longer producible
Here is the part I have not seen stated anywhere, and it is the single most important thing for a brand considering a listing.
Consumers on these platforms were trained on prices that were achievable only under the old structure. Duty-free entry, no customs entry, no destination inventory carry, no domestic warehousing, no local returns handling. Those prices set an expectation, and expectations move much more slowly than cost structures.
The platform's merchandising systems were also calibrated on those prices. Search ranking, category placement, promotional eligibility, the price bands that qualify for traffic: all of it was tuned against a cost base that no longer exists for anyone. When you list, you are evaluated against that calibration.
You are being invited to compete on a price surface built by an arbitrage, after the arbitrage has been removed, using a cost structure that never had it in the first place.
This is why the standard advice, list a few SKUs and see how it goes, produces the results it produces. The SKUs get no traffic at a defensible price, the seller discounts to buy visibility, the discounted price becomes the reference, and six months later the brand concludes that Temu does not work for premium products. What actually happened is that they entered a price competition against a historical cost structure and lost.
The leakage problem, and why it is worse than it was
The second-order risk is not that Temu shoppers think less of you. It is that your Temu price becomes the market's reference price for your product, and that reference then appears everywhere your customers evaluate you.
This risk used to be bounded. Price comparison required deliberate effort, and most customers did not make it. That bound has been removed in the last two years. Shopping assistants inside AI products now assemble price comparisons across marketplaces automatically, in the flow of a normal question, for a customer who never intended to comparison shop. Retail media and shopping feeds do the same thing on the search side.
The practical effect is that your price surface has collapsed. There is no longer a meaningful distinction between the price you publish on a marketplace and the price you publish on your own site, because both are inputs to the same comparison, and the comparison is being run on your behalf without your involvement. If you are thinking about how those systems form a view of your brand at all, the AI search and citation work is the adjacent problem.
Two rules follow, and they are more restrictive than most commercial teams will like.
Never list your hero SKU. The product that anchors your brand, carries your margin, and shows up in your own advertising should not have a second published price anywhere you do not control the merchandising. This is not snobbery, it is arithmetic: the hero SKU is the one most likely to be compared.
Only list SKUs you can defend as genuinely different. Different size, different pack configuration, different colourway, different bundle. Not different for the sake of it, different enough that a customer comparing the two would understand why the prices differ. If your answer to why is cheaper is because it is on Temu, you have not solved the problem, you have published it.
The cost of not listing, which is also real
The argument above is not an argument for abstaining, and I want to be careful here because the abstain case is usually made badly and then adopted for the wrong reasons.
Two genuine costs to staying off.
Discovery for people who do not search
A meaningful share of these platforms' users are not conducting product searches at all. They are browsing a feed, the way people browse short video, and buying what surfaces. That audience is not reachable by search advertising, is expensive to reach on social, and in many categories is younger and earlier in their category relationship than your existing customer base.
If your category is one where the first purchase is cheap and the lifetime value comes later, that audience is worth something, and there is no other efficient way to buy access to it.
The barrier to entry has fallen, which cuts both ways
Because both platforms have moved toward local fulfilment, listing no longer requires a China-based supply chain, a China entity, or acceptance of multi-week delivery. A US or EU brand with existing domestic warehousing can list from the inventory it already holds. Onboarding is fast by marketplace standards.
That lowers your cost of experimenting. It also lowers your competitors' cost of experimenting, which means the counterfactual to you not listing is not that your category stays absent from the platform. It is that your category appears there represented entirely by other people.
The compliance transfer nobody negotiated
The EUR 200 million DSA fine and the French enforcement actions are usually read by brand teams as reputational data points. Read them instead as an operating forecast.
When a platform is fined for failing to identify illegal products among its listings, the platform does not solve that problem by inspecting more products itself. It solves it by pushing documentation requirements onto sellers. Product traceability, compliance files, responsible-person details, test reports, ingredient and material disclosures, all demanded on the platform's timeline with delisting as the enforcement mechanism.
"A regulator fines the platform. The platform passes the obligation to the seller. The seller is you. That is the whole transmission mechanism, and it runs in one direction."
If you are an established brand with a compliance function, this is an advantage rather than a burden, and it is the most underrated reason to consider listing. The requirements that are painful for a thousand anonymous sellers are documents you already have. Enforcement waves that clear the category of low-quality competition are structurally good for you.
If you do not have those documents ready, do not list yet. Getting delisted mid-season for a missing compliance file is a worse outcome than never having listed, because it happens after you have built inventory against the channel.
A decision framework that actually resolves
Four questions, in order. If any of the first three come back the wrong way, stop.
Do we have a defensible SKU that is not our hero product and is genuinely differentiated from what we sell elsewhere? If no, the answer is no, and it stays no until product development creates one.
Do we have the compliance documentation for the destination market, in hand, today? If no, the answer is not yet, and the work is compliance work rather than commercial work.
Can we fulfil domestically from inventory we already hold? If no, you are taking on a working capital position to test a channel, which is the wrong order of operations.
What is the price we can hold, and are we prepared to be invisible at that price for the first quarter? If the honest answer is that we will discount to get traffic, do not start. The discount is the whole risk.
For brands that pass all four, treat the first two quarters as a measurement exercise with a fixed budget and a pre-agreed exit, and measure the thing that matters rather than the thing that is easy. The easy metric is channel revenue. The metric that matters is what happened to conversion rate and average selling price on your own site in the same period. If your direct channel softened by more than the marketplace contributed, you have moved revenue between pockets and paid a commission for the privilege.
If your broader question is how to sequence marketplaces against your own site at all, the Shopify versus Amazon FBA decision covers the same trade-off in a more familiar setting, and the logic transfers.
Frequently asked questions
Should Western brands sell on Temu?
Only with SKUs that are genuinely differentiated from what they sell elsewhere, never with a hero product, and only if destination-market compliance documentation is already in hand. The decisive risk is not brand perception, which is recoverable, but the creation of a published reference price that AI shopping assistants and comparison systems will apply to your other channels. Brands that cannot hold a defensible price on the platform should not start.
Has the end of de minimis actually hurt Temu and Shein?
Both restructured toward destination-market inventory and local sellers in response, which removed the cost advantage that underpinned their original pricing. Shein's Hong Kong listing disclosures show revenue growth slowing to 1.1 percent in the first quarter of 2026 alongside a $99 million net loss, after $41.8 billion of revenue and 8 percent growth in 2025, and a target valuation well below its earlier private mark. That is consistent with a business absorbing the loss of a structural advantage.
What is the semi-managed or local fulfilment model?
The seller holds inventory inside the destination country, sets more of its own pricing, and handles domestic fulfilment, while the platform provides the storefront, traffic and checkout. It replaced the China-direct parcel model as customs rules tightened, and it is why a Western brand can now list without any China supply chain at all.
How risky is this from a regulatory standpoint?
The regulatory pressure is on the platforms, but it transmits to sellers as documentation and traceability requirements enforced by delisting. The European Commission fined Temu EUR 200 million under the Digital Services Act in May 2026 over illegal-product risk assessment, and French authorities have imposed penalties on Shein reported to total more than EUR 210 million. Expect the requirements passed down to you to tighten rather than relax.
What should we measure in a pilot?
Not marketplace revenue. Measure conversion rate and average selling price on your own direct channel over the same period, against a pre-pilot baseline. Marketplace revenue that arrives alongside a decline in direct-channel selling price is revenue you moved rather than revenue you made, and you paid a commission on it.
Is listing reversible if it goes badly?
The listing is. The price is less so. Delisting removes the storefront but not the record: comparison engines, cached listings, screenshots and shopping assistants retain the price you published for a long time. Plan the entry price as though it were permanent, because in the systems that matter it approximately is.
The judgement
The interesting thing about Temu and Shein in 2026 is that they have become ordinary. Very large, very well funded, very good at traffic, and no longer operating on a structural advantage that Western brands could not replicate. The parcel arbitrage is gone. The regulatory tolerance is gone. What remains is a marketplace, which is a thing your commercial team already knows how to evaluate.
What has not become ordinary is the price expectation they created. That is the asset they still hold and the liability they hand to anyone who lists. A consumer who learned in 2023 what a dress or a phone case or a set of storage bins should cost did not unlearn it when the customs rules changed, and no seller on the platform, including the platforms themselves, can now produce that price profitably.
"They taught a hundred million people a price that nobody can charge any more. Joining the platform means being measured against it."
That is survivable, but only deliberately: a differentiated SKU, a price you will hold through an invisible quarter, compliance documents that are already in the drawer, and a measurement plan that watches your own channel rather than theirs. Every brand I have seen do badly here did the opposite of at least three of those, and most of them did the opposite of all four.
Share:
Free · 24 pages
Read Chapter 1 of the Playbook
AI Foundation Models — how DeepSeek and Qwen crossed a billion downloads while Western boards were counting Mag7 capex, and which layers of the AI stack the rotation has not reached. No excerpt tricks: it is the full chapter.
You will also get the weekly newsletter. Unsubscribe any time — the link is in every email.
The Book · $99
China Market Entry Playbook 2026
456 pages across 22 industry chapters — AI, semiconductors, EVs, payments, logistics, beauty, F&B, healthcare, pets, toys. Every chapter runs the same five questions: the anchor number, the rotation underneath it, which layers of the global stack it has reached, and what to do about it. English and Simplified Chinese editions included.
Ex-Transsion Global Digital Marketing Director · Effie Awards Greater China jury member · Guest lecturer, Peking University School of Economics · Writes Annie Chan Talk, your insider lens on China.
