Zalando, a prominent player in Europe’s online fashion industry, has introduced new tools leveraging artificial intelligence to improve customer interactions and provide enhanced insights into emerging trends. The Zalando Assistant offers a conversational interface for personalized shopping recommendations, while Trend Spotter delivers weekly updates on style trends from ten major European cities, including Stockholm.
Fabio Baum, General Manager for the Nordics & Benelux, discusses the development and objectives behind these tools. With extensive experience in strategy and growth, Baum provides insights into how these features align with Zalando’s broader efforts to meet changing consumer expectations and adapt to advancements in digital retail technologies.
Can you tell us about Zalando’s new AI-driven assistant and how it enhances the shopping experience for users across the Nordics?
– With one of the most relevant and extensive assortments in the market, Zalando recognizes that customers sometimes need extra guidance to find the perfect item, especially for specific occasions. While filters are useful, they can sometimes feel limiting when navigating a broad selection.
– That’s where the Zalando Assistant steps in. It empowers customers to navigate our assortment using intuitive, conversational queries like, “What should I wear to my friend’s wedding in December in Stockholm?” The assistant factors in contextual details such as location, weather, and occasion to provide personalized and relevant recommendations, making the shopping experience more seamless and inspiring.
How does the AI assistant utilize OpenAI’s language models and Zalando’s proprietary technology to provide personalized recommendations?
– We share our customers’ questions – and it’s important to note, only the questions – with OpenAI. Their advanced language models interpret the query, understanding precisely what the customer is looking for, and relay that information back to our system. From there, our proprietary technology processes the insights and suggests fashion items that match with the criteria of the request.
– Currently, the Zalando Assistant doesn’t access individual customer profiles or preferences. Instead, the personalization is driven entirely by the customer’s specific query. This makes it a highly tailored experience, offering real-time solutions that meet customer needs with precise and context-aware recommendations.
Could you explain how the Trend Spotter feature works and what insights it offers customers about style trends in different European cities?
– The Trend Spotter provides customers with a weekly snapshot of what’s trending in 10 major European fashion capitals, including Stockholm. Powered by Zalando’s proprietary data, it offers insights into emerging trends across these cities, giving customers a unique opportunity to explore the pulse of European fashion. The Trend Spotter isn’t based on sales, so not on what people are actually buying already. It’s all about people’s desire to own an item, so we use a combination of different factors, such as products being searched for, liked or added to people’s carts.
– For Stockholm, the feature reflects the city’s distinct blend of minimalist sophistication and functional design, showcasing how local preferences align with or diverge from broader global trends. Each featured item comes with an explanation of why it’s trending that week, whether it reflects a global movement or a uniquely Nordic style statement.
The Trend Spotter currently shares trends from the following cities: Amsterdam, Antwerp, Berlin, Copenhagen, London, Milan, Paris, Stockholm, Warsaw and Zurich.
What specific customer needs or behaviors prompted Zalando to develop and expand the AI assistant and Trend Spotter?
– We understand that customers appreciate a little extra guidance when navigating our extensive assortment. But beyond that, we’ve identified a broader shift in customer expectations, particularly among younger demographics. For example, 70% of Gen Z shoppers make purchase decisions while seeking inspiration, and 72% of this inspiration happens online.
– While we’ve consistently prioritized convenience – such as our recent improvement in Sweden, where most items are now delivered within 1-2 workdays – we know customers today expect even more. They seek personalized and inspiring shopping experiences. Features like the Zalando Assistant and Trend Spotter are designed to meet these evolving needs, providing tailored recommendations and trend insights that make shopping on Zalando both seamless and engaging.
In what ways does Zalando’s AI assistant address local trends and preferences? How does it customize recommendations based on factors like location, weather, and occasion?
– Currently, the Zalando Assistant’s recommendations are entirely shaped by the details provided in the customer’s question. For example, if you’re searching for an outfit for your best friend’s wedding in Stockholm in December, the assistant uses contextual factors like location, weather, and occasion, interpreted through advanced language models, to provide tailored suggestions.
– The accuracy of its recommendations improves as more specific information is given, much like interacting with a store assistant in real life. If you walk into a store and ask for dresses without specifying the length or preferred colors, the suggestions might not fully match your expectations. Similarly, the more details you share with the Zalando Assistant, the more precise and relevant the results become.
What kind of feedback have you received from customers since launching these AI-driven tools? Are there any surprising insights or adjustments you’ve made as a result?
– Customers have embraced our new AI-driven tools enthusiastically. Since its launch in the spring of 2023, over 1 million customers have interacted with the Zalando Assistant, with an average of four back-and-forth exchanges per session. Interestingly, customer queries are three times longer than those typically entered in the normal search bar, highlighting that Zalando Assistant isn’t replacing traditional search. Instead, it’s being used for complex, conversational queries, exactly how we intended. To me, this confirms that the assistant is fulfilling its role as a personalized advisor.
– Feedback has also guided key improvements. For example, in its early version, the assistant struggled with context when referring back to specific items. If a customer said, “I like the second item you showed me. Can you show it in different colors?”, the assistant sometimes misinterpreted which item was being referenced. Based on customer feedback, we have refined this functionality, ensuring smoother and more intuitive interactions.
From a broader industry perspective, how do you see AI transforming online fashion retail in the coming years? What role do you envision for Zalando in this transformation?
– Looking at the history of our industry, the first technological breakthrough was e-commerce, followed by mobile. I’m sure many in the industry still remember how we went from creating features that were mobile-responsive to focusing on mobile-first experiences. Today, we see the rise of generative artificial intelligence as the next major disruptor. One that will transform our industry completely, comparable to the development of the internet in the 1990s, if not greater.
– At Zalando, technology is deeply embedded in our DNA. We’ve leveraged AI and Machine Learning for years to enhance personalization, but generative AI unlocks entirely new possibilities. It enables us to offer even more tailored, engaging, and intuitive shopping experiences that cater to individual customer needs.
– As we continue to explore the potential of GenAI, we’re focused on learning from our customers’ feedback and finding ways to refine and elevate their experience. Our goal is to remain at the forefront of this transformation, ensuring Zalando not only adapts to but also helps shape the future of online fashion retail.
Looking ahead, are there any plans to further develop the AI assistant or expand its capabilities to include other features that enhance customer engagement?
– Looking ahead, we have several ideas for how we can further develop the Zalando Assistant to make it even more helpful for our customers. Currently, it isn’t connected to customer accounts, which means it doesn’t know individual preferences, such as our customers’ liked brands, your sizes or saved items. Building this connection is something we’re definitely considering, as it would allow us to personalize results not only based on the specific query but also by taking into account customer preferences and order history.
