The intention algorithm: how AI predicts our desires and what we will buy
Emma McFerran, CEO of Lyst – an aggregator of 27,000 brands and shops worldwide – explains how, in the age of artificial intelligence, ten million everyday actions can be transformed into an assessment of tastes and trends.
At the height of Facebook’s popularity, experts claimed that the algorithm would be capable of sensing a user’s intention to change their life even before they were aware of it themselves, by picking up on moods and subtle shifts in language. Today, some are delving even deeper, mapping contemporary desire through ten million daily actions. A search, a saved item, a purchase: it is this mass of signals that Lyst works with – a search engine and global aggregator that brings together 27,000 brands and shops and uses artificial intelligence to guide and personalise the multi-brand shopping experience. But what it observes goes beyond purchases: it traces the journey through which taste is formed, fragmented and changes direction.
Emma McFerran, aged 40, CEO since 2022 and with the company since its inception, emphasises that, to achieve this, it is necessary to combine the research aspect with the emotional one. At Lyst, they describe themselves as fashion architects, because the project stems from the intersection of art and science. ‘Users’ activity is constantly visible on the platform. We can see what people love and search for, what they add to their wishlists, and what they interact with,’ she explains. And she concludes with an almost programmatic definition of what they have built: “I’d call it a ‘platform of desire’. Lyst does not directly own the stock, which remains in the hands of retailers and brands: it organises choice and discovery through data.” The aggregator was founded in London in 2010, born of an insight that other sectors – from travel to music – had already experimented with: bringing together a vast range of offerings in one place and relying on data to make it accessible and personalised. McFerran joined when the company had just over ten people working out of a warehouse in East London – early enough for its founder, Chris Morton, to end up regarding her almost as a co-founder. Soon, observations of user behaviour gave rise to the Lyst Index, the quarterly ranking of the hottest fashion brands and products, fuelled by every search, save and purchase made by customers. The inspiration came not from the catwalks, but from music charts. ‘We took our cue from Billboard, applying to fashion the idea that there could be a sort of neutral benchmark capable of showing, in real time, where customers’ interest was concentrated and how it changed over time.’
Turning desire into data, however, means reaching the point where data alone is no longer enough. At Lyst, they call them silent criteria, the silent criteria: everything that influences a choice without ever being explicitly stated or typed into the search bar, such as the occasion, the social context or insecurity. McFerran gives the example of an invitation to a work cocktail party hosted by one’s partner: ‘A situation where you don’t know anyone, but you want to make a good impression and get the dress code right. You want to be comfortable yet chic; to stand out, but not too much. These nuances linger in your mind when you simply type in ‘black cocktail dress’, losing sight of all that crucial context,’ she explains.
It is in the space between those three words and what precedes them in the mind that artificial intelligence opens up a new field of research, attempting to get closer to the unspoken part of a question. It is more difficult to capture the opposite: “‘I hate it’, ‘I don’t like it’, ‘I’d never wear it’: in fashion, these reactions can be very visceral, but it’s tricky to pick up on them,’ he admits. Whilst the system is learning an ever-changing aesthetic vocabulary, the context is also shifting. ‘In the past, aspirations were much more top-down and uniform. Today, they have become complex and fragmented,’ he explains. Knowledge has also become more widely accessible: you can follow an entire creative team, track down an archive piece from the other side of the world, or get lost in online discussions about the texture of a fabric, the cut, or the quality. ‘The range of what we consider to be luxury, or what we can dream of, has expanded enormously,’ notes McFerran.
Of course, the big brands continue to attract a huge amount of attention, but they are no longer the only guiding principle. “We’re seeing that aesthetics, rather than the brand, are becoming the organising principle behind building a wardrobe. The questions now are: ‘What is the aesthetic? What is my style tribe? How do I put it together? What look am I going for?’ The old search formula of ‘brand plus keyword’ is being replaced by detailed descriptions of a visual world to be reconstructed. This attitude is particularly evident among Gen Z, for whom fashion and culture seem to move at almost the same pace. In the second quarter of 2026, the hottest item on the Lyst Index was the New York Knicks cap: searches rose by 4,276 per cent compared with the previous period. An item that becomes the symbol of a collective moment; McFerran interprets this as “the connection between a real-life moment – perhaps even a physical event they didn’t attend – and the demand for the fashion surrounding it”.
