Goddiva debuts AI virtual try-on to cut fashion returns

Goddiva has launched an AI-powered virtual try-on feature that lets shoppers upload an image and see themselves wearing dresses across its online store.

The London-based occasionwear retailer said the feature is initially available to a selected group of VIP customers. Goddiva plans a broader rollout in early summer.

The company said the tool appears on collection pages and individual product pages. It generates a visualisation of the customer in selected styles after the shopper uploads their own image.

Goddiva said the system does not rely on a simple overlay or face-based visual. It analyses body structure, proportions and size data. It then simulates how a garment fits, drapes and sits on the customer's body shape.

Returns pressure

Online fashion retailers have faced persistent issues with fit uncertainty and returns. Goddiva positioned the virtual try-on as a response to those pressures, with a focus on fit confidence and clearer expectations before purchase.

"This is about confidence and clarity," said Yathu Kanagaratnam, Head of Technology & AI Strategy, Goddiva.

Kanagaratnam described the visualisation as a direct representation of the shopper rather than a model or avatar.

"When customers can see themselves in a dress - not a model, not an avatar, but them - they can make better decisions. That means fewer returns, less frustration, and a far better overall experience," said Kanagaratnam.

Goddiva also linked the feature to the operational cost and logistics burden associated with high return rates in eCommerce. Retailers typically face additional shipping movements, repackaging work and processing time when customers send items back.

Sustainability angle

The company said it expects the feature to reduce its environmental impact by cutting the volume of returns. Goddiva pointed to emissions associated with repeat shipping and the waste created by repackaging.

Goddiva said the virtual try-on is part of its sustainability strategy, with fit accuracy positioned as a route to lower returns.

Technology stack

Goddiva said it built the platform in-house. It uses Google Gemini alongside proprietary AI systems developed by the retailer's technology team.

Goddiva said the feature works on an opt-in basis and meets GDPR requirements. The company said it processes customer images solely for the virtual try-on experience.

The retailer's focus on privacy reflects wider consumer sensitivity around the use of personal images in AI-based shopping tools. Fashion companies deploying similar features have faced scrutiny over how they collect, store and process customer data.

Next features

Goddiva also outlined further work tied to the same product roadmap. The company said it is developing an AI-powered size predictor.

The size predictor uses uploaded images and body analysis to recommend a suitable size. Goddiva said the feature remains in development and will arrive in a later phase.

Goddiva also said it has AI video generation in production. The company said the feature will show how fabrics move and catch the light on a customer's digital likeness in motion.

Goddiva said the video generation feature is scheduled to be available in summer 2026.

Broader push

The retailer said the virtual try-on extends an existing set of AI-based shopping and styling tools it introduced last year. Goddiva framed the work as part of a wider effort to refine product discovery and fit confidence on its platform.

Goddiva sells occasionwear for weddings, parties and special events. It operates online and positions its range around trend-led designs and accessible pricing.

"This is another major step in how we reimagine online fashion," said Kanagaratnam. "It's about reducing disappointment, respecting the customer's time, and building a smarter, more responsible way to shop for occasionwear."