Artificial intelligence, new consumers and future of fashion business

After two days walking the crowded aisles of eShow at Ifema, speaking with exhibitors, listening to speakers and reading flyers, one thing became clear: in 2026, using artificial intelligence (AI) no longer sets a company apart. The good news is that knowing what to use it for, where it adds value and where its usefulness ends can make it a valuable tool for your business and teams.

eShow was one of the seven events that Tech Show Madrid brought together on November 4 and 5 at the exhibition centre. The organisers expected around 30,000 trade visitors and over 500 companies, despite coinciding with the Barcelona Deep Tech Summit. More than 150 gathered in the e-commerce and digital marketing area. During the presentations, several fashion brands participated to explain how they apply AI in their offices and behind the scenes of their e-commerce operations.

From Coosy to Zalando, including Lady Pipa and Loewe, all the fashion companies that attended eShow already use AI, each on its own scale and with different objectives. None presented a replicable formula, as there is none. What they did share was a specific order of priority.

First come the brand, the product and the customer; then comes the technology that serves them.

Technology has been democratised, brand has not

Before adding more tools, some brands propose returning to the essentials. Carlota Posada Belbel, e-commerce manager at Spanish womenswear label Nicoli, opened her presentation by proposing “a digital reset to put the product back at the centre”. Her approach involves simplifying the user experience on the website and making it consistent with the brick and mortar store in aspects such as real-time stock and returns.

Coosy shares this approach and sums it up in one sentence. “Technology has been democratised. The brand has not”. Virginia Pozo Castro, founder and CEO of this Spanish womenswear brand, and Betty Lepina, its senior e-commerce and omnichannel expert, opened their talk with this idea. Coosy was founded in 2011 as a multi-brand online store and soon became a brand in its own right. By 2015, it already had 11 high street stores. Today it has 25 and is present in over 200 multi-brand points-of-sale across more than 30 countries. It launches around 400 styles per season with a team that its managers describe as “tiny”.

For Coosy, AI is “just” the latest in a series of technologies that have transformed the sector and allow brands to maximise the online channel. Having a website or an e-commerce platform no longer differentiates a brand, although failing to execute them properly can penalise it. A slow website or a poor mobile experience is now a handicap, not an option. The difference lies in knowing how to use technology with good judgement and building a recognisable brand identity. With the conversation increasingly taking place outside a brand's own website, brand imagery, product and the in-store experience are gaining weight.

Shopping habits are changing with the integration of large language models (LLMs) — such as ChatGPT, Claude or Gemini — into consumers' search and decision-making processes, as shoppers increasingly buy by asking questions. Instead of merely searching for products, they formulate specific, contextualised requests, such as “I want a dress for my friend's afternoon wedding in Madrid in October”. Coosy wants to be among the brands that appear in that answer. To achieve this, they explained, it is up to them to offer sufficient and relevant information on their website and across their other channels so that AI can identify and recommend the brand. Their advice for getting started is simple: ask the AI itself about the brand. Does it mention it? How does it describe it? What does it highlight? The answers help detect what information is missing from owned channels and which aspects should be reinforced.

From convincing consumer to convincing AI

Positioning brands within AI is undoubtedly one of the sector's main concerns. Quique Díaz Paradelo, CEO of Provoca and an expert in communication, marketing and public relations, addressed this as marketing director of Grupo Mascaró, owner of Pretty Ballerinas. His starting point is that while chatbots already help with searches, we are approaching a scenario where AI agents begin to decide and pay for purchases on behalf of the user. While the challenge until now was to convince the consumer, brands must now also convince the AI.

The first step is technical. The website must be open to AI crawlers, which many brands unknowingly block via the robots.txt file, the server or the content delivery network (CDN). It must also offer consistent product data — catalogue, price, sizes, availability, materials, manufacturing and care instructions. Paradelo recommends thoroughly developing the “frequently asked questions” section, as this is where AI finds data-backed answers and examples to resolve the doubts customers raise in their chats before purchasing.

However, the greatest weight lies outside owned channels. According to the data he presented, 84 percent of what influences AI comes from earned sources, in a context where the media, especially through lists of new arrivals and trends, is regaining prominence as an independent endorsement. This is followed by experts, forums and user reviews. Wikipedia is regaining weight according to his sources, while YouTube and LinkedIn are overtaking Instagram and TikTok as reference networks.

To measure a brand's position, he proposes a test that consists of asking the AI what it knows about the brand. This involves asking questions that name it, open-ended questions about the category and others that compare it with competitors. The results should be turned into a plan to correct erroneous data at the source, reinforce size guides and FAQs, work on media coverage and repeat the test. At Pretty Ballerinas, whose sales grew by at least 40 percent over the last year according to Paradelo, the exercise detected that Gemini did not identify its manufacturing in Spain, and that negative reviews on Trustpilot posed a risk they are now working to address.

Zalando takes its first steps in agentic commerce

German online fashion marketplace Zalando is working in the same direction on a different scale. Eloisa Siclari, general manager southern Europe, explained how the platform is taking its first steps in agentic commerce. “We cannot create a filter on our e-commerce site that says ‘December wedding in the Sierra de Madrid’, yet that is precisely what customers are asking AI”.

Zalando has enabled an assistant that allows shoppers to interact with the store in this way, taking into account what the customer has bought and returned, their usual price range or what they have left in their basket. According to Siclari, the assistant has received over 22 million visits so far in 2026. 40 percent of conversations begin with an automatic suggestion. The user asks an average of two questions, yet spends three times longer on the site than during a traditional search. Around 4 percent of conversations end with a product in the basket.

Prioritising when data and short-term demands overwhelm

Data, alongside AI and AI agents, completed the trio of the most repeated concepts during the event. Data is currently one of the most valuable assets of any company, as it allows businesses to know the customer better and make operations more efficient. However, “the help of AI must go hand in hand with human input, otherwise the essence is lost”, argued Teresa López, head of customer service at Spanish streetwear brand Scuffers. The label, like its current competitors in the segment, builds its narrative around community, a concept born from what is generated offline within a highly digital environment.

With more information than ever and an increasingly short-term outlook, how does one distinguish the urgent from the important? This question was the focus of a panel moderated by Cristina López Pancorbo from department store group El Corte Inglés, featuring a representative from Loewe, the Spanish luxury house owned by the LVMH group, alongside professionals from other sectors. The speakers agreed that this is not a linear decision, nor does it consist of arranging good ideas into a list of priorities, as new developments emerge every day that could completely change their course. “There are good ideas that have to wait because both time and people are limited, and the multiple technological solutions available are a bottomless pit”.

The key, according to the panel, is to maintain focus and work with a clear pattern, defined priorities and protocols to which every decision can be adjusted. Only from that starting point does it make sense to assess the impact of a new development on the strategy and decide whether to commit to it. Simply looking at data is not enough — and even less so looking at it to make it say what one wants to hear — as fashion remains a subjective industry driven by desire.

“You cannot gamble your reputation on a quick decision”. In the luxury sector where Loewe operates, customer trust is a particularly sensitive asset. The house uses AI intensively to ensure an impeccable experience in logistics and website navigation. However, someone who spends at least 3,000 euros on a handbag, or 100 euros on a tomato leaf-scented candle, does not want to be assisted by a bot on the phone. Brands must be clear about who their customer is and their expectations at each touchpoint, deciding where technology adds value and where human interaction is non-negotiable.

In a context where technology allows images to be generated with a prompt, Loewe celebrated its 180th anniversary with a watercolour animation, painted entirely by hand and directed by artist Isabel Garrett. Luxury houses Prada and Dior have followed suit with handcrafted Christmas campaigns, moving completely away from generative AI.

What value does AI add to brand communication?

Images and videos created with AI have gained significant ground on social media, the e-commerce sites of major brands and even within small businesses. For a small or medium-sized brand, producing content using these tools is now within easy reach. The question is not how much AI facilitates image production, but what value it adds to a brand's communication. Bárbara Morera, CEO of Envezdeflores, an online store for personalised gifts, championed personalisation as a competitive advantage. The company was founded to offer an alternative to more generic gifts, adding a name or specific details chosen for an individual. According to Morera, her customers value giving something practical and personal that does not look like a last-minute purchase. In this model, AI is used to show the same product with different names, initials or colours, something difficult to photograph piece by piece. These images coexist on their social media channels alongside more organic content featuring real people and product videos.

Lowering production costs does not always require AI. Rocío Gómez Buhigas, CEO of Spanish occasionwear brand Lady Pipa, encouraged making the most of available resources. “Every summer the Lady Pipa office fills up with international female interns who help us segment our social media presence”. AI can translate a message, yet it cannot capture the nuances of each culture. The company presented its US bridesmaid campaign as an example, photographed in Madrid's Retiro park with “five young Dutch women”. According to Buhigas, they fitted the aesthetic sought for that market, without the cost of a US influencer or travelling there. Even so, the brand has signed agreements with several sororities to generate organic content, as its strategy involves adapting content as much as possible to each country as it continues to grow internationally.

A team of seven or eight people creates content almost daily, and the company also invests in paid advertising and influencers to support its expansion phase. Its goal is to close 2026 with a turnover of 17.5 million euros, up from 11.3 million euros in 2025, and to make the leap into brick and mortar retail with pop-up stores in London, Amsterdam and Paris between March and May 2027.

Trust is won or lost in logistics and service

Valencian costume jewellery brand Singularu, which has 102 stores and plans to close the year with 110 after eight openings in less than two months, explained how it uses AI to tackle the logistical challenges of this growth. Its product is seemingly simple to transport, yet holds high emotional value as it is usually a gift linked to a special moment. “If the gift does not arrive on time for the customer who has trusted your one-hour in-store collection or express delivery, you have to be able to look after your relationship and reputation”, explained Daniel Capilla Sánchez, its head of logistics.

It is in this area that they see the greatest potential for AI, which allows them to inform the customer without relying on the logistics partner and to keep inventory updated and interconnected.

For an omnichannel industry like fashion, this shift adds complexity. Until recently, retail played out on two levels: online and offline. AI adds a third that evolves at a pace that not even the sector, accustomed to moving fast, can keep up with. Ana Gordo, head of e-commerce Europe at COS, part of the H&M group, summed it up by saying that the list of “how to succeed in business just grows and grows”. On that list, the brands that attended eShow agree on prioritising what no tool can replicate: their product, their customer and their identity.