An online fashion giant is extending its shop window into other retailers' websites. For fashion and beauty brands, the interesting question is not whether a robot can buy a handbag. It is who gets the customer relationship when it does.
THE DIGITAL DOLLHOUSE | INDEPENDENT EDITORIAL ANALYSIS | 3 NOVEMBER 2026 | FASHION, E-COMMERCE & CREATIVE TECHNOLOGY
Imagine the most important sale of your week. Your new collection is beautifully photographed. The product page is persuasive. Inventory is correct. The customer is ready to buy. There is one small problem: the customer never visits your website.
That future moved a little closer on 3 November 2026. Zalando introduced Zalando Extended, a planned way for customers to discover products from third-party online shops inside the Zalando app, even when those products are not already part of Zalando's assortment. Its first pilot partner is premium homeware retailer Westwing. An agent-assisted purchase and checkout experience is planned from 2027, not generally live today (Zalando, 2026; Reuters, 2026).
For retailers, this is a potentially useful new route to customers. For independent brands, it is also an invitation to ask an unfashionably practical question: if somebody else controls discovery, recommendation and checkout, which parts of the customer relationship do you still own?
What Zalando actually announced (and what it didn't)
The announcement arrived alongside Zalando's third-quarter financial results. Group revenue rose 5.1% year on year to €3.17 billion, while gross merchandise volume, the value of goods traded on the platform, grew 7.5% to €4.54 billion. The company reported 63.2 million active customers across its apps, a rise of 2.9% (Zalando, 2026). These are Zalando's reported business figures, not projections for brands that might join Extended.
The less glamorous numbers deserve equal attention. Zalando said its business-to-business revenue grew almost 28% to €354 million, while its retail-media revenue increased nearly 18% year on year. Partners accounted for 33.6% of consumer gross merchandise value, up 2.9 percentage points (Zalando, 2026). In other words, the company is increasingly making money by enabling other sellers, not only by stocking and selling products itself.
The company says its Extended pilot with Westwing began in 2026, starting with home and living. The AI-assisted checkout that allows a shopper to complete a purchase from Westwing without leaving Zalando is described as a 2027 ambition. The reported facts do not establish that every fashion retailer can sign up today, that every independent brand will get this distribution, or that these transactions are already converting at scale.
Reuters also reported a less flattering figure: Zalando's gross profit margin fell 0.9 percentage points to 38.7% during the quarter, with targeted discounts contributing to pressure, particularly in sneakers (Reuters, 2026). Growth, it turns out, can sit next to an uncomfortable margin conversation.
The new luxury is being found without surrendering your identity
There is a seductive pitch in any large marketplace: reach customers already prepared to shop, without first persuading them to find your tiny corner of the internet. A platform with tens of millions of customers is not easily recreated by a clever Instagram reel.
But reach is not the same as recognition. Suppose a shopper asks for a beautifully made burgundy work bag under £250. An intermediary could narrow the choice based on price, colour, availability, delivery and product attributes. Your brand may make the shortlist because its data is exceptionally clear. Equally, a shopper may remember the platform more vividly than the name stitched inside the bag.
That is not a prophecy about how Zalando's eventual recommendation system will work. It is a commercial scenario worth planning for as retailers expand their assortments and introduce conversational shopping tools. Product data helps products qualify for consideration. Distinctive brand ideas help people care which product is chosen.
The answer is not to plaster every pack shot with logos. It is to build identity into the product proposition itself: a recognisable silhouette, a useful point of view, truthful materials information, an unmistakable styling language, thoughtful packaging and a reason to recommend the item to a friend by name.
Four questions every UK fashion brand should ask
1. Does your product catalogue work when your homepage disappears?
For a multi-channel brand, each product needs a reliable identity outside the full-screen campaign film. Audit names, variant sizes, shade and colour attributes, material composition, fit guidance, stock status, delivery promises, returns, care instructions and photography. Keep descriptions consistent across your own website, retail partners and marketplaces.
Test the catalogue with a deliberately boring brief: black leather ankle boots, UK size 6, low heel, suitable for commuting. Could someone identify the right product without ever landing on your homepage? If not, the problem is probably information architecture, not a lack of fashionable adjectives.
For beauty brands, substitute shade, finish, ingredients, application instructions and relevant cautions. Especially in regulated categories, commercial convenience does not remove the obligation to make accurate, substantiated product claims.
2. Who owns the customer conversation after checkout?
Before signing an integration, ask who issues the order confirmation, handles delivery questions, owns the service relationship and may lawfully contact the buyer later. Separate the answers for the marketplace, seller, logistics provider and payment processor. Do not assume a platform's reach gives your brand unrestricted access to shopper data.
For UK businesses, personal information must be handled under applicable data-protection law. The ICO's data-sharing guidance emphasises that organisations receiving or sharing personal data need to understand their responsibilities, rather than treating a database as a free marketing asset (ICO, n.d.). The exact arrangement depends on the parties' roles and the agreed flow of information; take appropriate legal advice when needed.
Plan for a positive relationship the customer chooses, not one you quietly inherit. Exceptional packaging, clear care guidance and useful service can create a reason to seek out the brand again without trying to capture data improperly.
3. Can you measure profitable orders, not just new visibility?
A platform can report impressions, product views, attributable sales and orders. Your commercial dashboard needs more: contribution after product cost, commission, fulfilment, returns, discounts, media fees and customer service. A fast-growing retail-media line on the platform's income statement is a reminder that distribution and advertising can become separate expenses.
Build a channel-level model before allocating spend. As an illustrative example only, if a £100 product costs £36 to make, £10 to fulfil and return-provision, £18 in commissions and £12 in advertising, the remaining £24 is contribution before other overheads and taxes. Change those assumptions to your real figures. A marketplace order is not automatically a profitable one; neither is a direct sale if acquisition costs are high.
Set a test threshold in advance, compare like-for-like periods and audit incrementality where possible. If the sale would have arrived directly anyway, the distributor may have collected a fee for an order you were already likely to win.
4. Are you building brand memory or merely renting shelf space?
Retail media can work brilliantly when it captures genuine purchase intent. It can also tempt teams to optimise everything for the next click and forget why the customer should recognise the brand next season. Build campaigns that travel beyond a product tile: coherent visual codes, a point of view on style, credible creators, useful editorial and excellent product storytelling.
The strategic division of labour matters. Marketplace data should help people find the product. Brand strategy should make it memorable. Your own website should provide the deeper world: collections, craftsmanship, founder perspective, fit advice, and an experience worthy of an intentional visit.
What this means for beauty brands, not just fashion retailers
Zalando's headline announcement begins with homeware rather than beauty. That matters. There is no verified claim here that Zalando Extended is currently rolling out agent-led beauty checkout. Yet the underlying design challenge already exists across beauty e-commerce: how can a skin tint, serum or perfume remain recognisable when retailers arrange products into highly comparable grids?
Beauty founders should test descriptive accuracy before chasing distribution. Make shade ranges and textures clear. Explain product application without unsupported treatment claims. Give shoppers reliable ingredients and compatibility information where relevant, and ensure packaging and online photography depict the product faithfully.
For fragrance, where a screen cannot reproduce scent, discovery kits, sampling and good sensory storytelling may still do more commercial work than a flashy recommendation widget. Technology can reduce friction. It cannot replace a truthful, distinctive proposition.
The 30-day 'platform-ready, brand-distinctive' audit
Week one: catalogue hygiene. Select your ten most commercially important products. Check their attributes, imagery, availability, price consistency, delivery terms and returns information across all active channels. Record every discrepancy.
Week two: journey mapping. Walk through a marketplace order and a direct order as a customer. Document who handles support, how returns work, what first-party information each party legitimately receives and where the brand experience weakens.
Week three: creative distinction. Show your product images, stripped of brand marks, to a small group unfamiliar with the range. Ask what is recognisable, what is confusing and what value they can actually identify. This is a qualitative diagnostic, not a statistical consumer study.
Week four: measurement. Agree a contribution-margin view by channel and a sensible test for customer lifetime value and repeat purchase, using data you legitimately hold. Decide which channels deserve expansion and which need redesign before more budget.
A small British fashion label is not required to join every new marketplace, build an autonomous purchasing agent or surrender the front page of its brand. It is required to know why customers choose it, make that choice easy and understand the economics of reaching them.
A better question than 'should we use AI?'
The headline may involve a shopping agent, but the boardroom question is older: who introduces you to the customer, who persuades them, who completes the sale and who earns the loyalty afterwards?
Zalando's 3 November announcement makes that division of responsibility newly visible. It is a serious commercial signal, not proof of a fully functioning universal checkout or a forecast of inevitable brand disintermediation. For a founder, the useful response is neither panic nor breathless adoption. It is stronger product information, stronger creative identity, more deliberate distribution and more demanding measurement.
The shop window can move. Your point of view shouldn't.
At The Digital Dollhouse, we are a Knutsford-based digital marketing and creative technology agency combining brand strategy, creative advertising, SEO, ecommerce design and bespoke digital experiences. Explore our work or start a conversation about a more distinctive, commercially useful customer journey.
For a related but separate topic, read our Journal analysis of Google's changing shopping search. That article examines discovery on search engines; this one examines retail platforms, ownership of checkout and channel economics.
Frequently asked questions
What is Zalando Extended?
Zalando Extended is a platform initiative announced on 3 November 2026 to show products from other online shops within Zalando. Its first pilot partner is Westwing. AI-agent-assisted checkout inside the app is planned for 2027, rather than a broadly available service today (Zalando, 2026).
Should independent UK fashion brands join more marketplaces?
Not automatically. Decide based on relevant shoppers, commercial terms, fulfilment capacity, brand fit, contribution margin and the effect on existing channels. A marketplace should provide incremental value, not merely an extra dashboard.
What can a brand improve before agent-assisted shopping becomes common?
Start with product attributes and accuracy, distinctive photography and descriptions, consistent claims, clear returns and delivery information, and measurement of channel profitability. Those are sound ecommerce fundamentals regardless of how shopping technology evolves.
References
Information Commissioner's Office (ICO) (n.d.) Sharing personal data in databases and lists. Available at: https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/data-sharing/data-sharing-a-code-of-practice/sharing-personal-data-in-databases-and-lists/ (Accessed: 3 November 2026).
Reuters (2026) 'Retailer Zalando launches AI shopping agent as margins squeezed', 3 November. Available at: https://www.reuters.com/business/retail-consumer/retailer-zalando-beats-quarterly-sales-expectations-it-taps-ai-agents-2026-11-03/ (Accessed: 3 November 2026).
Zalando SE (2026) 'Zalando grows profitably in Q3, driven by very strong B2B expansion, and continues innovating through AI with Zalando Extended', 3 November. Available at: https://corporate.zalando.com/en/financials/zalando-q3-2026-results (Accessed: 3 November 2026).