In a world of almost limitless choice when it comes to fashion the big question for many shoppers is how do you choose easily and how do you choose well?
Many of us like the idea of appearing on trend, but are less happy about spending hours trawling through online stores or the high street to make this happen.
This is exactly where London based technology company, Thread, are making their name as an online platform for the style hungry and time poor modern man.
Now you may find it interesting that we have referred to what could simply be called a fashion company as a technology company.
The truth is that Thread is in fact a neat combination of both fashion acumen and technological substance.
Using a mix of human stylists and algorithms the Thread platform curates outfits for you, sourcing and recommending products which fit your personal parameters.
Every week your stylist fills your home page with clothes and outfits he or she thinks you’ll like based upon your selected fashion preferences, e.g. skinny or bootcut jeans.
These recommendations improve from your individual responses. By liking or disliking items, you help Thread’s algorithms learn your preferences, resulting in highly targeted suggestions from over 1900 brands.
This innovative use of machine learning enables each human stylist to deal with over 10,000 clients and has proved extremely successful, as Thread has now grown to help nearly 500,000 clients and now employs a team of 37 staff.
Thread have used Bytemark’s dedicated servers from the beginning to help power their platform, giving them the speed and resilience required to deal with thousands of customers on a daily basis.
We asked Ben to tell us a bit more about the company and some of the technical decisions he made whilst developing Thread’s infrastructure:
How did you validate the idea behind Thread?
Our initial version of Thread involved us walking around Covent Garden in London with a clipboard and a camera. We asked people if they wanted some styling advice from a top stylist for free and took their photo and email address. A stylist friend of ours then chose some clothes available online for each person and we emailed them the links.
While it was zero-tech (ok, we used email), it gave us a ton of information on how much interaction people wanted from a stylist online as well as a ton of feedback on whether there was any actual value in our idea.
Why do you think people love the service so much?
The same reason Thread exists - because most men have the same problem Kieran, Ben and I had.
I saw lots of people in the media, film stars, pop stars, as well as some friends, who had great style and an ability to look great effortlessly, and I could never manage to emulate this from shopping online or on the high street.
Men have a right to feel confident by looking great when they leave the house and there are far too many of us who can’t do this on our own.
What are your future ambitions?
We’re aiming to make Thread the default destination for men to get their clothes and styling advice. Navigating the high street and the myriad of online shops takes up too much time and is far too confusing for guys. To do this we’re working hard every day on improving the product.
Further down the line we’d love to build a similar service for women, as well as expand outside the UK.
What factors influenced your decision to use high-powered dedicated servers over cloud servers at Bytemark?
There are two main reasons: first, I’ve always liked to have more control over my network. While much improved in recent years, cloud servers often have inconsistent performance and it’s harder to determine why performance is suffering.
Secondly, Bytemark’s customer support is second to none. We’ve had to use their urgent support a couple of times now and it’s been impeccable.
Do you have a particular development philosophy?
I’m a firm believer in keeping things as simple as possible. While I get excited by new technologies and spend time building things at home to keep myself sharp, I’d rather use tried and tested software and ideas when it comes to the important things.
We have as few moving parts as possible. My favourite piece of software is PostgreSQL, it’s incredibly solid, reliable, powerful and just gets on with its job without falling over ─ I think parts of your stack should just get on with their work with minimal maintenance.
“I love using Thread to find specific items such as jeans. Usually it takes just 10 minutes to find something using their recommendations - it’s the best way to shop!”
– Sam Beckham, Web Developer at Bytemark