How YOOX NET-A-PORTER uses AI for product content, merchandising and clienteling
From the 2018 YooxMirror avatar styling app to a virtual stylist and group-wide recommendation, search and copy tools: what YNAP and parent LuxExperience have published about AI.
KEY TAKEAWAYS Summary by the editors
- In December 2018, Yoox launched YooxMirror, an iOS styling app in which deep learning networks extract product attributes from images to propose items that go together, and a 3D avatar named Daisy models outfits.
- In June 2020, YNAP announced an AI research project with academics at the University of Modena and Reggio Emilia, starting with visual search and virtual fitting, drawing on roughly twenty years of collected data.
- YNAP has also been reported to have developed a virtual stylist AI tool intended to complement its personal shoppers, learning from past stylists' work and customer purchase history, likes and dislikes.
- According to Glossy (2026), LuxExperience, the parent of Mytheresa, Net-a-Porter, Mr Porter and Yoox, says it uses algorithms for recommendations, search, merchandising, and product copy and imagery, with many features running on a Google Vertex AI partnership begun in 2021.
- Published sources give no measured results for these tools and the group's ownership changed, so YNAP-era claims should be read as historic.
YOOX NET-A-PORTER (YNAP) has experimented with AI since at least 2016, with documented projects in visual styling, visual search and a virtual stylist for personal shoppers. Today its brands sit inside LuxExperience, which reports wider use of algorithms for recommendations, search, merchandising and product copy and imagery. Evidence on outcomes is limited to announcements and interviews.
How has YNAP used AI for styling and clienteling?
On 10 December 2018, FashionUnited reported that Yoox had introduced YooxMirror, described as an AI-powered virtual styling suite aimed at millennials, designed by Yoox's research and development team and available on its iOS app. It showed more than 250 products a week across nine fashion themes, with a curated selection refreshed daily, and let users swipe through products and see personalised recommendations that complement what they are viewing.
A 3D avatar named Daisy modelled outfits and guided users, who could mix and match clothing and accessories, share looks, save them to a wishlist or add them to the basket. According to the report, algorithms detect visual elements such as colour, pattern and shape in product images, deep learning networks extract product attributes to pick items that go well together, and virtual reality technology fits the items onto the 3D model.
Separately, a Springwise report describes a virtual stylist AI tool meant to complement YNAP's personal shoppers. It learns from past stylists' work and customers' purchase histories, likes and dislikes to suggest items that go well together. The article carries no publication date; it mentions only that the company had experimented with AI since 2016 and had recently opened a new tech hub in London.
What did the Modena research project involve?
On 2 June 2020, The Industry Fashion reported that YNAP launched an AI research project with academics at the University of Modena and Reggio Emilia. Its stated aims were to build cutting-edge technological solutions that tailor the online shopping experience to individual shoppers, beginning with visual search and virtual fitting technology, and to use roughly twenty years of collected data, including tens of millions of images.
The value of that data point is instructive for merchandising teams: a long, well-structured image and product archive is a prerequisite for visual models. The article describes aims, not results.
| Date | Initiative | Area | What was reported |
|---|---|---|---|
| December 2018 | YooxMirror | Styling, personalisation | Deep learning on product images, 3D avatar, iOS app |
| Undated (post tech hub opening) | Virtual stylist | Clienteling | Tool to complement personal shoppers, trained on stylist work and customer history |
| June 2020 | University of Modena and Reggio Emilia project | Visual search, virtual fitting | Research aims using about twenty years of data |
| 2021 onward | Google Vertex AI partnership | Search, recommendations | Many customer-facing features run on it, per Glossy (2026) |
| 2026 reporting | Group algorithms | Merchandising, copy, imagery | Applications listed, no results quantified |
How does LuxExperience use AI for product content and merchandising?
Glossy reported in May 2026, following LuxExperience's fiscal third-quarter earnings, that CEO Michael Kliger said the company had expanded widely its use of algorithms to improve the customer experience. The reported applications are on-site and newsletter product recommendations, on-site search, on-site merchandising, and product copy and imagery. Kliger also cited large benefits in software development during the technology transformation of Net-a-Porter and Mr Porter since the acquisition.
Many customer-facing features run on a Google Vertex AI partnership that began in 2021. These tools use product data and user events such as clicks, purchases and add-to-cart actions to return ranked product results for individual users. The article gives no figures attributing revenue or margin gains to AI and does not name specific Net-a-Porter or Mr Porter features beyond that list.
What does the YNAP case show and not show?
- Show: a long record of experimentation, from avatar-based styling to research on visual search.
- Show: AI positioned as support for human stylists, not a replacement, in the virtual stylist description.
- Show: dependence on product data quality, including image archives and event data.
- Do not show: measured uplift in conversion, returns or content productivity.
- Do not show: how much product content is now generated by AI, or what review steps exist.
Readers should also note the change of ownership: the group's brands are now reported under LuxExperience, and parts of this record relate to the earlier YNAP organisation. Statements from 2018 to 2020 describe past projects and may not reflect present operations.
What data supports AI in luxury merchandising?
The sources highlight three data assets. The first is a large image archive: the 2020 research project cites roughly twenty years of collected data including tens of millions of images, which is what visual search and virtual fitting models need. The second is stylist knowledge: the virtual stylist tool learns from past stylists' work, which means human expertise becomes training data. The third is behavioural data such as clicks, purchases and add-to-cart events, which the 2026 report says feed ranking tools.
Each asset needs maintenance. Images need consistent attributes, stylist decisions need to be recorded in a structured way, and event data needs to be clean and lawfully collected. The sources describe the existence of these assets, not how they were governed.
What should retailers check before using AI for product content?
Generated copy and imagery raise questions about accuracy, brand voice, intellectual property and disclosure. A description that misstates a material or care instruction can cause returns or compliance problems, and generated images of products may mislead if they differ from the real item. Retailers should define which content types may be generated, who reviews them and how corrections flow back. None of this is described for YNAP or LuxExperience in the sources reviewed.
Frequently asked questions
How does YOOX NET-A-PORTER use AI?
Published sources describe an AI avatar styling app (YooxMirror, 2018), a virtual stylist tool to support personal shoppers, a 2020 research project on visual search and virtual fitting, and, via LuxExperience, algorithms for recommendations, search, merchandising, and product copy and imagery.
What is YooxMirror?
YooxMirror is an iOS styling app launched by Yoox in December 2018. It used deep learning on product images to suggest items that go well together and a 3D avatar named Daisy to show outfits. FashionUnited reported on its launch.
Does Net-a-Porter use generative AI?
Glossy reported in 2026 that LuxExperience uses algorithms for product copy and imagery, which indicates generative tools, but it does not name specific Net-a-Porter features, volumes or review processes.
Has YNAP published results from its AI projects?
Not in the sources reviewed. Announcements describe features and aims, and the 2026 Glossy report explicitly gives no figures attributing revenue or margin gains to AI.
One edition every weekday morning. Read in five minutes. Free for industry professionals.
SOURCES
- FashionUnited: Yoox unveils AI-powered virtual styling suite
- The Industry Fashion: YNAP launches AI research project to transform online shopping experience
- Springwise: AI tool complements work of stylists
- Glossy: Luxury Briefing, Ralph Lauren and LuxExperience turn to AI for styling, search and high-value shoppers
