Fashion's AI future

Harnessing the seemingly unlimited powers of artificial intelligence could be key to the industry’s success in the next decade

Data as much as design now drives the fashion industry. E-commerce – though still less significant in pure sales terms than bricks-and-mortar stores – generates an astonishing amount of information about what we buy, where we buy it, or what we looked at before we bought it (even the 75 per cent of possible sales left abandoned in digital shopping carts is a useful indicator of intent).

The industry has decided that these huge streams of data can be crunched and sifted so that consumers can be more precisely targeted, teased and tempted. Data analytics is the new engine of desire, and artificial intelligence (AI) and machine learning promise pattern-hunting and prediction making of a far higher order. AI is also, of course, threatening to automate away all manner of professions. The creative industries, fashion included, are not immune to the threat. McKinsey’s ‘The State of Fashion’ 2018 report suggested that 20 to 30 per cent of jobs currently done by fashion designers could eventually be automated.

Pulling in and analysing sales and search data, as well as smartly scanning social media with new image recognition software, AI can create a kind of super-charged mood board, pumping out algorithm-informed choices for colour, cuts, fabrics, patterns and so on. It marks a fundamental shift in the relationship between the brand design team and the consumer. ‘The power of social media means trends are now more often established by consumers, as opposed to retailers and editors,’ says McKinsey’s 2019 fashion report. ‘In an era of fast-changing preferences, being able to respond to shifting demand, and tailor production accordingly, makes sense.

Fashion is seeing the start of a seismic shift where products are “pulled” into the market based on actual demand, rather than “pushed” based on best guesses and forecasts.’