Zalando

How Zalando is using AI and data science to become Europe’s most fashionable tech company

Technology is becoming central to the success of fashion and retail companies around the world. From the time when they used technology to mainly deliver their goods to technology being used extensively to decide their products, technology is everywhere. Zalando, the largest online-only fashion retailer in Europe, is going a step further.

At Zalando, co-founder and co-CEO Robert Gentz explains, technology is being leveraged to deliver shoppers a “more personalised experience” on the site as well as on the app. It is also using AI and data science to improve its product catalogue and increasingly, AI is becoming vital for Zalando to reach its sustainability goals.

AI to deliver a personalised user experience

The success and failure of an online fashion retailer not only boils down to choice but also the ability to bring those choices right to customer’s browsing. In an interview with Business of Fashion, Gentz says this is one of the challenges for Zalando, which has 1.4 million different items on its retail platform.

With over 48 million customers, Zalando is using artificial intelligence (AI) to bring the right merchandise to the right customer. Gentz further explains that the company is using an algorithmic fashion companion to solve its matchmaking problem. This companion works on the basis of items users have purchased in the past.

The easiest way to think of it is to look at your smartphone’s operating system. As you tend to use your phone more and more, the operating system becomes smart enough to not only recommend the right app but also recommend the time of the day since it understands which apps you use at what point during the day.

Similarly, the fashion companion designed by Zalando combines fitting items to create an outfit for each and every customer, which Gentz says is done using learning around how people combine items.

Is this algorithm working? Well, Gentz says the click-through rates and buythrough
rates are higher for the items produced using this algorithm. In a nutshell, Zalando is using algorithms that are “continuously improving with feedback loops from customer data as well as human feedback” produced internally.

AI is so deeply fused into its website and the app that it is not just limited to the outfits shown to you. Gentz says AI begins working the minute someone downloads the app or visits the site. This is usually in the form of options to express your favourite brands, your size, and other personal choices.

All of these user inputs form the basis of Zalando’s AI application to begin building a truly personalised user and shopping experience for its customers. “The Zalando shop looks different to every single customer once they actually have an interaction with us,” Gentz says.

Use of technology to curate and reduce returns

The AI program that Zalando has built is capable of understanding a customer’s past behaviour and offer suggestions that prove to be perfect. However, it is not all AI that does the curation. Gentz says shopping for an outfit is an “emotional experience” and hence the company employs “fashion people to help the technology people.”

Zalando calls itself Europe’s most fashionable tech company and to stay fashionable, it sees a world where physical changing rooms will not exist by 2030. Gentz says the world will reach a point where the experience will be the same everywhere leading to accurate sizing and options to make better choices.

At Zalando, the fashion retailer is driving this change by first analysing the items returned by its customers and taking feedback from them. With the help of these returns, Zalando is able to build a data graph and is able to further refine its recommendation. “We have already been able to reduce size-related returns by 10 per cent,” he says.

However, the coming attraction is one where these fashion retailers move towards whole-body measurements. Gentz sees use of 3D technology and body measurement technology to create an accurate profile of each and every customer.

Solving big data and algorithmic problems to become sustainable

Sustainability can be seen from a number of different lenses. One lens through which Gentz views and one that could be the most critical is logistics. He says the company has one of its biggest tech teams working on convenience and logistics. These teams are working to solve the challenge of placing an item in a warehouse.

The specific challenge being where to allocate an item so it has the greatest proximity to a customer across a warehouse network. By lowering delivery times and avoiding single item shipments, Gentz says Zalando will be able to drive sustainability.

In order to get there, the fashion retailer employs about 2,500 software engineers working across various teams. “When we have large-scale projects, we try to bring the different disciplines to the table and have them all looking at this problem,” Gentz explains.

Zalando is already using technology from platforms like Lalaland to make its shopping experience diverse and inclusive. In April, Zalando rolled out circular.fashion’s circular design criteria to brands but Gentz says sustainability is a data and collaboration problem.

The company is working with brands very early in the design process to ensure less resources are consumed and decisions are driven by data. Zalando sees use of quality data and increased collaboration as the most effective way to reduce overproduction or wrong production.

Access to quality data

Whether it’s customising product discovery or improving logistics or becoming sustainable, the most crucial piece of element in Zalando’s mission is data. However, we know that access to clean data or data offering valuable insights is not always easy. Gentz acknowledges these challenges and is tackling them in a smart way.

Firstly, he says they are not perfect at it but the company has set “ownerships for specific amounts of data” it produces. The company also constantly engages with the stakeholders on culture of data cleanliness and the need to get better data.

What to read next?

1200 800 Editorial Staff

Editorial Staff

My name is HAL 9000, how can I assist you?
This website uses cookies to ensure the best possible experience. By clicking accept, you agree to our use of cookies and similar technologies.
Privacy Policy