A product can look right until you start thinking about how you’ll actually use it. Will that jacket be warm enough for the school run? Can the bag fit a laptop as well as everything else? And what makes one option more suitable than another?

These are the kinds of questions I kept thinking about while reading Zalando’s latest update on its AI shopping assistant, published on November 11th. Much of the discussion about AI shopping focuses on finding products more easily. But I’m more curious about what happens after the first recommendation, when a customer needs to talk through whether something is right for them.

What Zalando is reporting

In its 3 November financial announcement, Zalando reports more than 22 million unique Assistant visitors this year.

The subsequent update describes availability across 29 markets in local languages, queries averaging three times the length of traditional searches and more than two questions per conversation. Suggested prompts start 40% of conversations. These are company-reported figures.

The publication does not disclose the query sample, measurement period or market breakdown. Longer conversations might indicate useful exploration or difficulty finding an answer; prompt-led starts also complicate interpreting them as spontaneous demand.

Zalando additionally reports a 4% add-to-cart increase from an A/B test prioritizing products available in customers’ sizes. It does not provide sample size, duration, statistical uncertainty or clarify whether the increase is relative. This is not evidence of increased purchase conversion. For me, the question is what an ecommerce team would need to examine to understand whether a conversation helped someone choose.

The next question asks more of product content

Imagine someone looking for a jacket for a rainy weekend away. They might start with a broad request, then add that they need something lightweight, without a particular material, and comfortable over a thick jumper. This is just an example, rather than a finding from Zalando’s data. But it helps explain the information challenge.

A description calling a jacket versatile or ideal for everyday adventures gives little basis for answering those questions. Composition, measurements, construction details and substantiated weather-protection information would offer something more useful.

The challenge also involves knowing where an answer ends. If a product specification does not mention a material, that omission cannot establish that the material is absent. If water resistance has not been verified, a recommendation should preserve that uncertainty. A useful review question for brands is whether each explanation of suitability can be traced back to a product fact.

Comparing products means explaining the differences

A customer may understand two products individually and still struggle to choose between them.

Take two bags. One listing gives external dimensions, while the other describes the laptop compartment. Both contain measurements, but they do not answer the same question. A comparison that treats them as equivalent could sound precise while leaving the customer with the wrong impression.

For ecommerce teams, this suggests reviewing similar products together. Can someone explain the meaningful differences using only the published information? Do attributes use consistent definitions? Is one product missing a detail that the other provides?

The same exercise can reveal gaps in sizing information. A size label alone may leave someone uncertain about fit. (I can tell you all about the struggle to find suitable leather gloves when sizes differ from brand to brand😊). Clear distinctions between body measurements and garment measurements, alongside relevant fit guidance, would give a comparison more context. FAQs can help with recurring questions, particularly where an answer needs explanation. But adding more questions to a page will not resolve an incomplete or inconsistent specification.

Local language needs local understanding

For brands working across markets, a useful test is whether the information supports the way people naturally describe their needs. A catalogue term may be accurate without being the term a customer would use. Someone might describe an occasion, a fit problem or a material they want to avoid. Their question may require clarification before a recommendation is possible.

Local reviewers can help by writing realistic questions about a product, then checking whether the available content supports an answer. Where does terminology need explanation? Which size convention is being used? Are market-specific conditions clear?

On-site search offers a useful starting point. In search-localisation work, I’ve found that product names are only part of the picture. Everyday expressions, synonyms and common misspellings also matter when considering how someone might look for a product. Conversational discovery takes that question further. Can the information support someone describing an occasion, a preference or something they want to avoid? Search terms can help connect the language, but the product information still needs to support the answer. The purpose is to preserve both the facts and their meaning. A fluent answer is of limited use if it quietly changes what the product can do.

What this means for ecommerce businesses

What I find useful about conversational shopping is how it brings familiar customer questions into the same exchange. Someone might move from discovering a product to comparing alternatives, checking fit and asking about delivery, adding context along the way.

For ecommerce teams, that creates a closer connection between information that may sit in different places: product descriptions, specifications, sizing guides, FAQs and delivery pages. Each contributes something to the customer’s decision. In a conversation, those details need to make sense together.

It also leaves room for clarification. Sometimes a recommendation depends on knowing a little more about what someone needs. Sometimes the available information supports a clear answer; sometimes there is still a detail to check.

That’s the part I keep coming back to. A customer’s first question may help them find a product. The questions that follow help them decide whether it’s right for them. As shopping becomes more conversational, those exchanges deserve a place in how we think about product information.

Sources

• Zalando, From conversational tool to a lifestyle agent: how the Zalando Assistant has grown up, 11 November 2026; live version checked 12 November 2026.
• Zalando, Q3 2026 results announcement, 3 November 2026.

————————————————

Written by Iusti Ikert, Founder of In Between the Lines — a messaging and customer experience alignment studio for scaling product-led brands. I work with growing DTC companies to align brand promise, marketing, and operational delivery so they can scale without friction.

Connect on LinkedIn or follow on Instagram for insights on sustainable growth, messaging clarity, and customer experience alignment.