Generative AI Revolutionizes Virtual Try-On: Will Shoppers Embrace It?

Zalando’s Virtual Try-On Journey: Transforming Online Retail through Generative AI

It’s been nearly eight years since Zalando, the German e-tailer, first began experimenting with virtual try-on technology. This innovative approach is revolutionizing how customers shop online, enabling them to visualize clothing on their own bodies before making a purchase—a game changer in the world of e-commerce.

The Road to Virtual Try-On: Milestones and Innovations

Back in 2018, Zalando took the initiative to develop proprietary software capable of generating images of customers adorned in items from its extensive catalogue. The pivotal moment arrived with the emergence of open-sourced large language models (LLMs), providing Zalando with an unparalleled boost in image-generation capabilities. This technological evolution set the stage for beta testing with small groups in late 2022. By 2023, Zalando began uncovering substantial benefits: a test carried out in April resulted in a remarkable 40% reduction in return rates, as shared by Reza Shirvany, Zalando’s director of applied science. The company is now gearing up to launch its virtual try-on feature for all customers by 2026.

How Generative AI Lowers Barriers to Innovation

Generative AI has transformed the landscape for many businesses, including Zalando, enabling them to innovate rapidly and effectively. As Shirvany noted, the traditional barriers to developing advanced technology in-house are diminishing due to the widespread investment in LLMs. This shift has greatly improved the realism of virtual try-on experiences, addressing a long-standing challenge for online retailers: accurately portraying how garments fit on real bodies.

The Impact of Generative AI on Realism

Historically, virtual try-on technologies often required users to scan their entire bodies or create avatars as rough approximations of their real-world selves. In contrast, advancements in generative AI enable users to simply upload selfies, instantly generating try-on images. This leap in technology allows for a more accurate representation of how clothing hangs and moves on diverse body types. Google has recognized this potential, introducing features that allow users to fit virtual products onto models that reflect a broad range of sizes—showing a commitment to inclusive fashion.

The Surge of AI-Powered Virtual Try-On Startups

With the growing interest in virtual try-on solutions, numerous startups like Doji, Zelig, and Stiled have emerged, each offering unique features aimed at enhancing user experiences. Major software companies are also investing heavily in this technology. For instance, Google’s recent features allow users to visualize themselves in clothes through uploaded photos, becoming a favorite tool among shoppers. According to Lilian Rincon, Google’s vice president of product management, virtual try-on has surged to the forefront of user preferences, garnering extensive shares across social media platforms.

Overcoming Adoption Challenges

Despite the buzz around virtual try-on technologies, mass adoption is still a challenge. A survey conducted by eMarketer revealed that only a mere 1.4% of adults regularly use virtual try-on features. To bridge this gap, software startups are focusing on improving user experience through faster image generation and by offering more engaging options, such as 3D imagery and videos that lend realism to the try-on process.

Evolution of Virtual Try-On as a Styling Tool

To drive consumer engagement, some startups are rebranding virtual try-on as a stylistic tool rather than a mere solution for returns. For instance, Doji, which launched in 2024, emphasizes its app’s ability to generate outfit recommendations and inspire users to explore various looks. Dorian Dargan, Doji’s co-founder, highlights the importance of establishing an emotional connection with brands through the experience.

Flexibility and Expansion in the E-Commerce Scene

In an effort to facilitate wider access to virtual try-on solutions, some providers are turning to browser extensions to reach more customers. Doji’s recent launch of an extension for Apple’s Safari enables users to overlay products onto their images from an expansive range of brands. This approach aims to create habitual usage of virtual try-on features, encouraging consumers to engage with the technology regularly.

Measuring Success and Proving Value

The future landscape of virtual try-on hinges on convincing retailers of its profitability. Zelig’s Sandy Sholl stressed the importance of partnerships and data management to attribute sales increases directly to the use of virtual try-on software. The overarching goal is to turn this promising technology into an essential revenue-generating tool for e-commerce businesses.

Moving Forward with Technological Innovations

As generative AI continues to revolutionize the retail landscape, the quality and fidelity of virtual try-on images will only improve. Investment in user-friendly features and compelling user experiences will be critical to fostering a new era of online shopping, where seeing is believing—and buying. The journey of Zalando and its contemporaries exemplifies how technology, when harnessed effectively, can reshape our understanding of fashion retail and consumer interaction.