Online fashion retailer Farfetch has been attracting attention this year.
In August, founder Jose Neves said a New York floatation was the "next logical stage" for the company, with a valuation rumored to be $5 billion, and in February it made headlines when Natalie Massenet joined as non-executive co-chair.
Massenet founded rival Net-A-Porter in 2000 and sold her stake in it to Richemont for an estimated £50 million ($66.8 million) in 2010, before it merged with Italian website Yoox in 2015.
And Farfetch's future success may be down to how it recommends products to shoppers. Chief Marketing Officer John Veichmanis said he takes inspiration from how Spotify recommends music to listeners when thinking about the best ways to promote the platform.
Speaking to CNBC, Veichmanis said he would like people to be able to discover up-and-coming designers via a similar type of algorithm. "Music is very similar, it is very personal. I love Spotify because it introduces me to loads of artists I would never have listened to and I suppose that's where I take our inspiration from," he said.
"If you look at music, someone can sit in their bedroom now, create an amazing track and distribute it themselves, and I think that has to come to the luxury (fashion) space as well."
Farfetch lets individual fashion boutiques sell to consumers, rather than holding stock itself, and the site has around 280,000 items for sale at any one time.
"We use information consumers have given us to tell them better stories (about products), and that's an emerging space. I don't think anyone is doing this very well and it's something we are really keen to learn," Veichmanis said.
Creativity and technology
Veichmanis, who has worked at various tech companies including Apple and Skype, said his job is made harder because people who shop for luxury fashion want something unique.
"A lot of the focus is on data science and looking at computer vision to understand which garments go together really well (or) what type of fabrics, rather than saying, 'Well somebody's bought this, and 10 other people have bought this other one.' That's completely the opposite of how people buy fashion, they want a unique perspective," he said.
To that end, Farfetch has more than 200 people in its marketing team, of which around 25 percent are in data science and analytics, bringing in a much-needed collaboration between creative people and tech experts, according to Veichmanis.
"Our perspective is that the industry hasn't really moved on: there is a lot of capability to target consumers, but then often you will talk to the creative teams telling the stories and they are completely disconnected from what the technology can do. What we are trying to do is put those people together," he told CNBC.
Net-A-Porter owes some of its success to its fashion content, including digital magazine the Edit and printed title Porter, and Veichmanis said it's also a key part of marketing for Farfetch, ahead of advertising. "We are publishing content every day, we are not thinking about paying for Q1 or Q3 (advertising). If we see something that is relevant, we publish it."
Farfetch also has about 100 personal shoppers for its biggest-spending customers and around 25 people who merchandise the website, putting products together to make outfits.
Machines are unlikely to be able to completely take over the role of curator, but the company spends much of its research and development budget in working out the best way to recommend products to people — and not just the top spenders.
Having such a large product selection might delight fashionistas, said Veichmanis, "but that's not useful unless you can make it relevant to a consumer."