From 2027, Zalando’s own AI agent will buy things from shops that are not on Zalando. The customer stays inside the Zalando app. The shop never integrates with Zalando’s catalogue, never changes its logistics, and never meets the customer.
Zalando calls it Zalando Extended. It was announced inside a quarterly results release, in a few paragraphs, with no executive quoted beside it. That is an odd place to put something this large.
What Zalando actually said
Products from other online shops, which Zalando does not sell itself, get added to the Zalando platform. Zalando’s own AI agent buys them on the customer’s behalf. In Zalando’s words, “customers will never have to leave the trusted Zalando app.”
For the shops, the pitch is that they reach Zalando’s customers and gain sales without complex technical integration or changes to their logistics, and that they can advertise through Zalando’s retail media business.
It starts with what Zalando calls complex categories, bulky home goods among them, and is meant to widen over time to products from shops across the internet. Westwing, which Zalando describes as Europe’s leading premium home and living destination, is the first pilot partner. The cooperation began in 2026, and from 2027 Zalando’s agent will check out Westwing’s products inside the Zalando app.
One thing to keep separate. The Zalando Assistant is a different product, a conversational search tool that more than 22 million unique visitors used this year. It helps a customer find things. Extended buys them. The two conflate easily and they are not the same announcement.
From the publisher
Everything we know about running Amazon, in one course.
The commercial strategy, taught by operators who run our own brands.
The bargain is inverted
To sell on a marketplace, normally, you integrate. A product feed, a category mapping, fee schedules, fulfilment terms, a seller agreement. The work is the price of the shelf space.
Extended removes all of it, and that is what makes it strange rather than merely convenient. The shop is not a seller on Zalando at all. Zalando’s agent buys from the shop’s own storefront, the way a customer would. The shop ships its own order on its own terms, and an order simply arrives.
That is not a lighter version of a marketplace listing. It is a different commercial relationship, and it comes with a different set of questions, none of which has been answered.
Tool comparison · FastMoss vs Kalodata
A channel with no published economics cannot be compared against the ones you already sell on, which leaves your own category data doing the work. FastMoss and Kalodata do that job for TikTok Shop, reporting category-level demand and product movement independently of any platform’s own account.
Comparison coming soon
FTC disclosure: our tool comparisons carry affiliate links. If you sign up through one we may earn a commission, at no extra cost to you. We rank on merit, never on commission, and the verdict is written by a human.
The four questions that decide whether this is good for you
Who is the merchant of record? If Zalando’s agent buys from a shop, has the shop sold to Zalando, or to the consumer with Zalando acting for them? That answer sets who issues the invoice, where the VAT lands, which consumer-protection rules apply and who the customer takes a complaint to. Nothing has been published.
What does it cost? No commission, no fee schedule, no margin structure, no retail media pricing. A channel with no published economics cannot be compared against the ones you already sell on.
Who owns the customer? The customer never leaves Zalando’s app, which means the shop receives an order and nothing else. No contact, no data, no second purchase. For a brand that has spent years building direct relationships, that is the real cost of the free integration, and it does not appear on any fee schedule.
Who absorbs the dispute? The shop’s own returns and after-sales terms will apply to a customer who believes they bought from Zalando. This lane saw the same shape at Shopify a week ago: an agent answers for a merchant, the answer is not quite the merchant’s policy, and the merchant absorbs the difference. Agents produce an answer either way. The only question is whose terms they produce it from.
Why Zalando is doing it
The same release carries the reason, though Zalando does not state one.
Zalando’s consumer business is growing and getting less profitable. B2C revenue rose 2.8 percent to 2,829.2 million euros, while B2C adjusted EBIT fell from 76.6 million euros to 64.4 million. Its business-to-business arm did the opposite: revenue up almost 28 percent to 354.4 million euros, adjusted EBIT more than doubling from 19.6 million to 42.2 million. B2B is about 11 percent of revenue and close to 40 percent of adjusted group EBIT.
Two more lines point the same way. The partner share of B2C gross merchandise value rose 2.9 percentage points to 33.6 percent. Retail media revenue grew nearly 18 percent.
Read together, the shape is clear enough. Selling things itself is where Zalando’s margin is thinning, and the money is in other people’s goods and other people’s advertising. Extended adds assortment and ad inventory without buying stock, building integrations or touching a warehouse. It is growth that costs almost nothing to carry.
That is a reading of figures in the release rather than anything Zalando has said. The only named quote in the whole announcement is the CFO on the ABOUT YOU integration.
Tool comparison · FastMoss vs Kalodata
When a platform’s margin moves from selling goods to carrying other people’s, the terms usually follow. Our head-to-head of FastMoss and Kalodata looks at how each tool reports category demand and product performance across European markets.
Comparison coming soon
FTC disclosure: our tool comparisons carry affiliate links. If you sign up through one we may earn a commission, at no extra cost to you. We rank on merit, never on commission, and the verdict is written by a human.
The part that is bigger than a pilot
Westwing is one partner in one category, and the agent does not transact until next year. On its own it is a small thing.
The sentence to notice is the ambition attached to it: products from shops across the internet. If that is where this goes, Zalando stops being a place where you list your products and becomes a layer that buys them from wherever they already are. The shelf is no longer something you apply for.
Whether a shop can decline, how a shop joins, and whether joining is required at all are not set out anywhere. Zalando says partner shops, which reads as consent. The stated ambition is wider than a partner programme, and the gap between those two is not explained.
What a seller does differently on Monday morning
Nothing this week. One named partner, one category, and no transactions until 2027.
- If you sell home, living or other bulky goods in Europe, put Zalando Extended on the list of channels to ask about in 2027 planning. It does not belong in a 2026 budget.
- Before joining anything like this, get three answers in writing: who the merchant of record is, what it costs, and who handles returns. Those three answers are the entire deal, and none of them is public.
- Read your own returns and after-sales terms as though a customer who never saw them is about to rely on them. That is what an agent-placed order does to a policy page.
- Do not read no integration required as no work required. An order with no customer attached to it is a different business from a marketplace listing, whatever it costs.
Where the honest uncertainty sits
Zalando has published no market scope for Extended. Which countries it covers is not stated anywhere, so a European seller cannot tell whether it reaches them.
No commercial terms exist in public. No fee, no commission, no merchant-of-record position, no returns or liability position, no retail media rates.
Westwing is the only named partner and the only date given is from 2027. There is no launch day, no second partner and no published list of categories beyond bulky home goods.
The explanation of why Zalando is building this is drawn from segment figures in the same release, not from anything Zalando said. Zalando gives no rationale, and attaches no executive to the announcement at all.
And products from shops across the internet is a stated ambition, not a published programme. How a shop joins, and whether a shop could find itself included without joining, is not addressed.
The three-line version
- If you run the channel: nothing until 2027. If you sell bulky home goods in Europe, put Extended on the 2027 planning list and get merchant of record, fees and returns in writing before anything else.
- If you have just taken on the channel: the thing to understand is that Extended is not a marketplace listing. Zalando’s agent buys from the shop’s own storefront, so the shop never integrates and never meets the customer.
- If you sign the budget: approve nothing. No fee, no liability position and no market list has been published, and the only partner named is a pilot that does not transact until next year.
Sources
- Zalando: https://corporate.zalando.com/en/financials/zalando-q3-2026-results
- Marketing4eCommerce: https://marketing4ecommerce.net/zalando-extended-ia-marketplace/