How do you cut through the boredom some men feel when shopping for clothes? For our Head of Marketing Andy Penfold, personal shopping site Thread.com has a few ideas…
I’ve got a new friend. Her name is Brooke and she may or may not be real. It’s a great relationship because she gives me advice, and wants me to be the very best version of myself that I can be.
Brooke is my virtual personal shopper on Thread.com.
Thread.com is an A.I.-assisted shopping site that is easily the smartest, slickest online shopping experience I’ve ever tried. Retailers both online and offline can learn some seriously useful lessons just by trying it out.
The site starts by asking you a series of questions about yourself. First, what brands do you tend to wear? This gives Brooke an idea of what brands’ clothes fit you. You’re then asked for vital statistics (height, weight etc) how you’d describe your build, and what sort of sums you tend to spend on shirts, trousers, jumpers and shoes.
Then the site shows you a load of pictures of blokes dressed in various ways. You simply click the ones you like. You then go through a series of screens that asks you what clothes you have that you like, and what you’re in the market for.
When it’s all done, Brooke recommends a load of clothes for you, and even constructs a personalised web page of ‘tips’ for you. For example, I learned that tall, skinny chaps shouldn’t tuck in their shirts, because doing so makes your legs appear even more gangly than they already look (obviously I’m paraphrasing – Brooke’s advice was much more tactful).
The whole experience, for me, was great, and the proof really is in the pudding. I blew a load of money on some new clothes.
So, what can Thread.com teach the world of retail? Here goes.
I’m a bloke who tends to fall out with my partner on shopping trips. I get bored almost as soon as I’ve parked the car. I don’t like anything in the shops, and if I do find something I think is alright, I don’t like the way it looks when I try it on.
As a result, I don’t like my wardrobe, because I can’t face updating it.
Thread.com seems to have identified me as a target segment – and crucially discovered I’m not alone. And given the wad of cash I threw at it on my first go using it, they got it bang on.
Thread.com used to serve women and men, but it’s currently just for men. The reason? The company’s CEO says that the women’s fashion market is completely different to men’s. Thread.com made an early decision to focus on one thing, and do it brilliantly.
That kind of focus requires digital product management that understands the business and the commercial imperatives behind it. Retailers looking to go into ecommerce in a big way will get nowhere without a good product manager.
Given the fact that walking past a coffee vendor in a shopping mall can cause my interest in clothes shopping to drop off a cliff, Thread’s see, click, and move on UX was exactly the experience I needed.
User experience is a fascinating discipline – the user rarely does what you expect. Optimising digital experiences (and experiences in the real world, for that matter) is a question of testing them on real people and learning from the data you collect.
Thread.com combines little personalised touches with the machine-learned algorithm that does the real work for Brooke and her ilk. For example, the site only shows you clothes in your size, removing the need to pick – personalising the user experience as well as the content.
But the big piece of personalisation work is done by ‘Thimble’ – Thread.com’s proprietary machine learning algorithm. Brooke’s involvement is to choose a range of outfits that might suit the user. Thimble then takes over – sifting through a catalogue of 200,000 items to find the actual items that might work for the user, based on the stylist’s outfit picks and the user’s data.
Using machine learning to augment human expertise in this way is a very current use of the technology (while we wait for Google and Apple et al to create robots communicative and tactful enough to be personal stylists).
As an online personal stylist Brooke relies on a vast suite of high-quality content that she can send to me. I get five recommendations a week – a combination of looks and actual items, curated into actual advice about when and how to wear, and why they suit my tall, skinny body and smart-casual work environment specifically.
Pictures look fantastic, copy is nicely done. And, because of the whole premise of the site, it’s all totally relevant to me.
Of course, not every retail brand can turn to sophisticated machine learning at the drop of a hat. But understanding how some of these agile start-up firms are playing with retail and disrupting it might just spark an idea that could change the way you do business with your customers.