Fashion search in the age of Generative AI 

The extended use of generative AI is fundamentally reshaping how consumers discover fashion products online, moving the industry away from traditional keyword-based ecommerce search towards a more conversational and personalised discovery experience. 

Traditional ecommerce and web search relied on users translating their needs into specific keywords, such as: 

• “black midi dress”  
• “running shoes men”  
• “best winter coat”  

Alternatively, LLM-powered search changes this dynamic by allowing consumers to describe intent conversationally: 

“I need a smart casual outfit for a work dinner in October.” 

Users are looking for recommendations, comparisons, product research and reviews e.g.  

• What’s the best luxury tote under £500?  

• Compare top running shoes 

Therefore, brands should be willing to change the way they are presenting their product information to adapt to this new way of searching, ensuring they remain visible and present consumers with personalised results to accelerate the shopping experience. 

Product discovery is evolving beyond traditional keyword-based search, as AI-powered platforms increasingly enable consumers to discover products through image uploads, social media references, voice interactions, and conversational assistants. In this context, multimodal discovery refers to AI systems that combine multiple forms of input and interaction formats including text, image, video, voice, and conversational context to better interpret consumer intent and deliver more relevant and personalised recommendations. 

This shift is particularly significant within the fashion industry, where the discovery journey is highly visual, emotional and inspirational. Consumers are no longer searching solely for products, but for aesthetics, occasions, and style identities that are often difficult to express through conventional search queries. Emerging platforms such as Daydream and Capsule represent this next generation of discovery systems, using machine learning, image recognition, and conversational AI to create more contextual, personalised, and intuitive shopping experiences. Behaving more as a stylish and personal shopper than a traditional search engine. 

Additionally, modern LLMs are supporting try-on features, especially in fashion, eyewear, makeup and accessories.

This is possible due to multimodal AI models being combined with technology such as AR (augmented reality) software, computer vision or body estimation, that allows users to upload a photo and ‘try on’ items digitally.

Whilst this technology is still being tested and presents some initial challenges around data protection or the accuracy of the technology issued, it’s likely to become more relevant and continue to develop. Fashion brands will benefit from reducing return rates due to sizing issues, increased conversion rates, and better customer insights.

Alongside the raise of LLMs, the SERP landscape is busier than ever with organic search listings, paid & shopping listings, image results and AIOs, so it’s not to anyone’s surprise that customers are relying more on generative AI to reduce their shopping research time.

AI Overviews traditionally appeared for long term informational queries; however, this is not the case anymore and their use for shopping queries has increased. Nowadays, 14% of shopping queries return AIOs.

Like LLMs, AIOs also help speed up the discovery journey, they present summaries, detailed recommendations, product comparisons as well as embedded shopping recommendations; all this information is presented before reaching organic listings.

In this context, generated AI Overviews are reducing CTRs across organic listings, meaning less traffic is driven through the organic channel. Some studies suggest that clicks to organic results are down by 38% due to AIOs.

Whilst clicks through organic listings may decline with the appearance of AIOs and the use of LLMs for product discovery, this AI generated technology presents numerous advantages for ecommerce businesses, but more so for fashion brands.

One of the biggest challenges for the retail industry is that online inventories are huge, presenting customers with almost infinite options, making their navigation journey overwhelming by the amount of choice available not only on traditional search, but also across social media platforms. Fashion brands have the added obstacle of trend cycles which means they are constantly expanding inventories.

The result is simple, poor user satisfaction that translates into abandoned purchases. A recent publication from Business of Fashion suggest that 74% of shoppers abandon purchases because of excessive choice.

Here’s where generative AI plays a key role in reducing decision fatigue through intelligent curation and personalised discovery. Moving away from manual filtering to get through thousands of SKUs before reaching the products consumers are looking for.

Today’s digital environment is defined by infinite inventory shrinking consumer attention spans, and so a competitive advantage not only depends on offering more products, but on helping consumers discover fewer, more relevant choices.

Additionally, modern LLMs are supporting try-on features, especially in fashion, eyewear, makeup and accessories.

This is possible due to multimodal AI models being combined with technology such as AR (augmented reality) software, computer vision or body estimation, that allows users to upload a photo and ‘try on’ items digitally.

Whilst this technology is still being tested and presents some initial challenges around data protection or the accuracy of the technology issued, it’s likely to become more relevant and continue to develop. Fashion brands will benefit from reducing return rates due to sizing issues, increased conversion rates, and better customer insights.

Big ecommerce platforms such as Zalando or Asos have already started integrating generative AI to improve their product discovery experience.

Zalando

Zalando, one of the largest ecommerce platforms for fashion, has developed an integrated shopping assistant in collaboration with OpenAI and powered by large language models; the assistant is embedded directly within the Zalando app and website, enabling users to ask questions such as what to wear for a specific event or how to achieve a particular aesthetic, and receive tailored outfit suggestions in real time.

The results speak for themselves; 25% [1] increase in product clicks as well as 40% products added to the wishlist.

Asos

Asos is working on reshaping their strategy, moving away from a high volume and promotional-led model to an inspirational shopping experience model.

They are increasingly using AI to create a functional reduction in choice even if the catalogue stays large by ranking products more intelligently, personalising feeds, prioritising “likely-to-convert” items, and suppressing irrelevant SKUs.

Their CEO recently stated: “We’re putting a lot of effort into improving the customer experience through virtual try-on, personalisation and ‘just for you’ features, continually introducing new tools to improve the customer experience. The goal is ultra-personalisation, so every time a customer visits Asos, it feels unique to them.” [2]

Showing highly relevant products reduces the noise and leads to faster shopping decisions, which will translate into higher sales.

Amazon

Amazon reported $43.2 [3] billion revenue in 2025 in the UK market alone, making this market their second largest European market after Germany.

This ecommerce giant has been pioneer in applying AI to improve the user journey, focusing their efforts in creating a seamless shopping experience, facilitating product discovery with highly personalise outcomes. They are combining different AI-powered solutions such as natural language search, visual search tools like StyleSnap and conversational AI via Alexa to deliver relevant product suggestions in real time. With this, they are managing to simplify product discovery and reduce fatigue, which in turn will translate into an increase in conversion rates, as well as higher basket spend through complementary product recommendations.

Amazon Fashion is not an exception and AI-driven technology is helping customers find outfits, styles and specific items faster.

The below is an example of how StyleSnap is helping users find similar products via image search:

[1] https://openai.com/index/zalando/

[2] https://www.just-style.com/news/ai-personalisation-in-discounts-out-in-growth-ambitions-says-asos/

[3] https://www.statista.com/statistics/1035592/net-sales-amazon-united-kingdom-uk/

Zara

Zara, the insignia brand of Inditex, is using AI to make shopping faster and provide a more interactive and personalised experience.

The fast fashion brand is not only using AI shopping assistants to help users discover products quicker using natural language models, images or even voice search. The brand launched their AI-powered “try-on” feature, which is currently available on their mobile app.

The “try-on” feature allows shoppers to upload photos, the system then generates a 3D avatar where users can preview and style an outfit before buying it; users can also mix and match items, rotate looks in motion, save styles and add products to the cart.

This technology is aimed to minimise returns and “bracketing” (when shoppers buy multiple versions of the same item intending to keep one and return the rest), encourage conversions, and provide an engaging and personalised shopping experience.

This shift toward AI-driven discovery is not limited to global giants; fashion brands of all sizes are beginning to measure and optimise their visibility within LLM-powered search and generative AI platforms. Understanding how brands like Mango or Desigual adapt and perform in these emerging AI-powered environments provides insight into the next generation of digital discovery, where relevance, context, and personalisation determine which products capture consumer attention.

Mango

Mango’s AI assistant works as a virtual stylist, helping users to find products based on natural language input. Rather than just showing a flat product listing, the assistant suggests complete outfits offering links to products in stock and available in the user’s region.

Like other platforms, Mango uses a multimodal system that allows users to discover products via image search, which helps bridge the gap between inspiration from other platforms such as social media and the shopping experience.

The below table shows Mango’s performance across LLMs vs some of the closest competitors.

Mango’s share of voice in the UK market is sitting just under 4%, where other retail brands such as Next or Zara are dominating this space – Next share of voice is nearly at 88%.

Source: Ahrefs

When it comes to AIOs, Mango’s rankings suggest an increase in terms returning AIO results, which is in line with recent findings about overviews appearing more for shopping searches.

In this case, AIO prevalence has increased by 233.5% YoY in May, with currently 6.4% of Mango’s ranking keywords in the top 20 positions returning AI Overviews. However, Mango is only present on 17.8% of these results, which presents a huge opportunity to increase their visibility on AI-powered searches.

Source: Ahrefs – Top 20 keyword rankings

Desigual

Desigual appear to also be integrating AI-driven technologies to help product discovery, however, it’s not as advanced as features we have seen across the big players.  Whilst long contextual searches appear to serve relevant results, their AI assistant chat can only be used for ‘order’ related actions e.g. returns, modifying delivery details etc. rather than for products discovery like the assistant on Mango’s website.

The below table shows Desigual’s performance across LLMs vs some of the closest competitors.

Desigual’s share of voice in the UK market is sitting at 0.15%, where other retail brands such as H&M are leading with an over 81% share of voice.

Source: Ahrefs

Desigual’s top 20 rankings have been following a declining trend over the past few months, however, like we’ve seen on Mango, the number of keywords that are now returning AIOs is increasing over time. Desigual has a strong presence on these terms (appearing on 64% of these rankings), however, most of these are branded terms.

There’s a huge opportunity for brands like Desigual to invest their efforts in increasing their visibility not only in traditional search, but also across ever growing AI-driven results and platforms such as AIOs or LLMs.

Source: Ahrefs – Top 20 keyword rankings

Whilst luxury brands may not be facing the same issues than large ecommerce sites such as Asos or Amazon with their huge inventories, they could benefit from the use of AI technology to provide a more personalised and engaging experience.

The below section outlines how some luxury fashion brands are using AI, as well as diving into their performance in LLMs & AIOs.

Loewe

Loewe follows a more traditional approach, with a well curated and editorial presentation supporting the luxury brand position. They don’t appear to make the most of AI-driven applications such as shopping assistant or natural language search; an example is their search feature not returning accurate results when using long-tail contextualise keywords:

Prada

Prada follows a similar approach than Loewe, focusing on showcasing products through high-quality imagery, videos, and runway content, reflecting a luxury, brand-first strategy.

Although their search functionality seems to be working better than Loewe’s for longer tail terms, their live chat doesn’t work as a shopping assistant. This approach is an interesting contrast when comparing with other brands such as Mango, or big ecommerce sites.

Gucci

On the other hand, Gucci is not only keeping their curated, editorial look, but it is also utilising AI to enhance customers experience positioning themselves as one of the most innovative brands in the luxury industry. Some of the things they are currently doing include:

• Chatbot that works as a shopping assistant, engaging and encouraging conversations with users, providing product recommendations and new products suggestions to complete an outfit.
• Virtual try on feature for shoes and watches on their mobile app

When it comes to performance across LLMs, we can see that Louis Vuitton received the most citations, followed by Gucci.

Interestingly, most citations for luxury brands are being driven by Grok, which is the LLM that pulls live insights, trends, and user sentiment directly from the X platform.

On the other hand, Balenciaga and Loewe present the lowest visibility on LLMs vs competitors

Source: Ahrefs

Below are some of the examples of prompts that are being used in ChatGPT and return designer brand mentions, showing that there’s big search demand for brands to capture.

Source: Ahrefs

Generative AI is rapidly reshaping the fashion discovery journey, transforming search from a traditional keyword-driven experience into a more conversational, personalised, and increasingly multimodal. As consumers rely more on AI-powered recommendations, visual search, and conversational assistants, visibility is no longer defined solely by traditional rankings, but by how effectively brands can deliver relevant, contextual, and curated experiences using AI-powered systems and platforms.

For fashion brands, this presents both a challenge and an opportunity. Brands that adapt their content, product data, and discovery experiences for AI-powered environments will be better positioned to remain visible, reduce friction in the customer journey, and drive stronger engagement and conversions. As generative AI continues to evolve, success will increasingly depend on combining technological innovation with seamless, personalised shopping experiences.

Get in touch with us today to discuss how we can support your business’s organic performance through our SEO services, content strategy and digital PR support.