Kieran O'Neill, CEO of London fashion startup Thread, is building a new way of shopping for clothes.
His site uses online stylists, as well as artificial inteligence (AI) and machine learning, to create a personalised way to shop.
But that can cause problems when he visits standard clothes shops.
"We went to Liberty for a bit and I went to the men's section," O'Neill said in an interview at Thread's London office. "[I] went to the rack and was browsing and I looked at it and it wasn't my size. I felt this rage, this offence that why would you bother showing it to me if it's not actually in my size?"
That's the problem that Thread is trying to solve. Walk around a clothes shop and chances are only a small proportion of the clothes on sale will both fit you and look good on you (not to mention fit your budget). But if you had a personal shopper with you, they could learn your style and size, taking you straight to the clothes that work best.
Thread is using artificial intelligence to help people buy clothes
Thread, which was founded in London in 2012 by O'Neill, CTO Ben Phillips, and creative director Ben Kucsan, pairs users with stylists who can provide shopping advice through the site.
Sign up to Thread and you'll be asked to provide some photographs of yourself, along with your measurements, what's currently in your wardrobe, and your budget. The company then uses that information and assigns you a virtual stylist who will start suggesting clothes for you to buy.
Obviously, Thread can't afford to hire personal shoppers to learn all they can about every single one of the site's users. So it uses a mixture of artificial intelligence and machine learning to help its stylists out.
"We have a pre-screen step where stylists go through and remove everything from our partners which they don't want to personally endorse," O'Neill explained to Business Insider.
Thread's stylists then look up what you want to buy, and they suggest individual items as well as full outfits. But after that initial human involvement, Thread uses its algorithms, known as "Thimble" internally, to do some of the heavy lifting.
Once a stylist has decided on an olive green T-shirt, for example, the algorithm looks to find the best olive green T-shirt for the customer. O'Neill said that would be "a really hard problem for a human to do" because of the volume of clothes to sort through.
"So that's a really good place for machine learning where you pull in lots of data from all the different partners you have. We have about 200,000 items from our partners. It's the combination of curation plus AI which has worked really well for us."
Another reason Thread uses AI is that it doesn't forget anything, unlike humans. "If you have a relationship with a stylist here for four years and you mentioned something four years ago that you liked or disliked, it’s likely the stylist would forget," said CTO Ben Phillips. "Whereas a computer never forgets."
A fashion site that just showed everyone exactly what they wanted all the time might not actually be the best idea. It's important to vary suggestions so that new trends develop and people discover new styles. Spotify, for example, proactively suggests new music that you haven't heard before.
Thread does something similar, said Phillips. "Every so often [we] send you something that's a bit not accurate or in an area of your data that we don't really have, so 'here's some skinny jeans.' We wouldn't blindly give those skinny jeans to everyone, it would be people that it would make sense, but maybe not an explicit thing that we think you would like and then we get feedback on that and that improves your personal data."
It wants to add new types of data
Why stop at knowing your budget or size? The more data that Thread has, the better. Phillips said that "weather is super interesting." Thread could learn upcoming weather conditions and use that in the text of the messages it sends to users, Phillips said.
O'Neill has another idea for data to bring into Thread: Your social media accounts. He mentioned Facebook, Pinterest, Instagram, and Twitter by name, speculating that if the data was properly sorted then it could help his site to send better recommendations. "You can maybe learn that someone is really into rock music," he suggested, "and you kind of bring in darker aesthetics into the recommendations."
Thead is also keen to add deep learning to its site so that its algorithms can understand what's actually in photos. "Say you wear baggy trousers," Phillips said. "Maybe you should think about slimming them down or something."
Thread wants to bring back women's clothing but 'there's no set date yet'
Thread used to sell women's clothes as well as men's clothing, but it eventually decided to focus on men's clothing only. "It became very clear after three or four months that it would be impossible to make something that was a breakthrough experience for both genders at once," O'Neill said.
"So we took the really tough decision to do just UK men for a while until we nailed the experience. We picked men because we started the business because we have this pain that we want to solve in our own lives and it would be inauthentic to do women because it was a bigger market, for example."
O'Neill explained that men and women shop in different ways. Women, for example, often shop for clothes for specific events, but men are more likely to make do with what's in their wardrobe. "You would design the experience slightly differently," he said.
"We'll definitely do womenswear, there's no set date yet. It's not something that we're actively working on right now, but I'm super keen to do it. We had at least 400, 500 women who were on the platform when we took them all out. And a lot of them are friends of mine who have bugged me."
Thread would 'love' to have physical stores
O'Neill said that the idea of Thread opening physical stores is "super interesting." "I think there's a really powerful thing about expressing the brand in a physical way that is just not possible online," he said. "I can see us having a few stores in London or in a major city. But I don't see us having like 500 stores in the UK or anything."
"I would love to work on it right now, it's just we're 35 people, there's only so much you can do at once. And our approach is to do a really small number of things, try and do it really well, rather than do 25 things all a bit average."
Join the conversation about this story »
NOW WATCH: Forget the iPhone 7 — here are 13 reasons the next iPhone will blow everyone away