Thread.com excels with personalisation at its core.
It’s not often that I have an ecommerce experience that inspires me to write about it, and certainly not one that’s as revolutionary as Thread.com.
Thread offers a personalised stylist service and boasts a wide range of high-street fashion brands. There have been a few personal shopper services over the years. Some use the subscription business model, others charge a fee. Others encourage try-before-you-buy by sending out styles and returning those which you don’t like. This is Stitchfix's unique selling point.
While Thread are a “small” UK fashion retailer, they have seven years under their belt, a reported 1 million customers (so I’m far from a pioneer), and no sign of slowing down with their most recent funding round last year totalled a tidy $22m from the H&M group.
Thread aren’t currently charging customers or merchants and, in addition to the sales, are providing valuable anonymised customer data to merchants. It’s an incredible service that provides a lot of value, so it’s not clear what the long-term business model direction is in order to sustain it. One thing Thread are betting on is that they can become a shopping destination, providing customers with choice and merchants with new customers. The question will be whether they can grow to be this destination more quickly than other merchants can acquire or integrate their own AI solutions through tools such as IntelliStyle, vue.ai, or Stylitics.
The personal shopper
The recommendations on Thread appear to be achieved through style combinations that are created by people, and then the machine learning chooses which of these styles to show to you.
While it is possible to shop without doing so, creating an account is heavily encouraged, and it’s quite an involved process. You are asked a series of questions to gauge your style and pricing preferences as well to gather your sizing information.
Once you’ve created the account, you are paired with a Thread stylist and your homepage is updated with some initial recommendations: “Styled by Millie, personalised by our Algorithms”. New style suggestions are then emailed out on a weekly basis.
Where the experience really excels is that once you’re logged in, there’s no other attention grabbing content on the homepage. It doesn’t push categories of products, or particular promotions, or recent blog posts. It takes a leaf out of our online lives and creates a feed of recommended styles.
This is in contrast to how most ecommerce sites typically do things, and certainly how most do personalisation. We love recommending Nosto, it’s a must-have tool, whether it’s recommending products or personalising content to the customer. However, we’ve always used tactically in certain positions and rarely, if ever, is the entire page given over to its recommendations.
One of the advantages of this approach is that it reduces the overload of information that you experience during a typical online fashion shop; paralysed by the time you get to page 4 of tan chinos. It’s a great way of combating the paradox of choice.
Another key to the success of these recommendations, is the way that styles have been paired with lifestyle images, creating a really nice “shoppable content” effect. Most shoppable content, whether it’s done with a tool like Olapic for your customers’ instagram posts, or Zmags for your own images, the use of image hotspots or popups can feel clunky. The prominence and pride of place that Threads gives its recommendations enables a much more pleasant experience that encourages use.
Building the customer relationship
There have been a few personal shopper services over the years. During the early stages, these would typically “do things that don’t scale” and the recommendations would be primarily people-driven.
In order to scale long-term, these recommendations would need to be more automated, and indeed Thread do boast a blend of AI and people.
So, after signup, I get a welcome message from my stylist, Millie. I haven’t taken her up on the offer of talking through my style choices to date, but it was a great way to kick off my interaction with the brand.
When browsing styles on your personal home page, you can provide simple feedback on each recommendation with a “I like this” or a “I don’t like this”. Regardless of whether I think that is probably just feeding the machine learning algorithms, it still felt like I was providing Millie with feedback.
Unfortunately, keeping this balance between AI and a real customer relationship is a challenge. While I was under no illusion that Millie was standing around waiting to provide me with recommendations, the day after I placed my order that I’d been working up to for a couple of weeks, I received an email from Millie with my next 5 product recommendations.
This was at odds with what a real relationship with a personal shopper would have been like at this point. I have to say I almost expected a follow-up email to my order telling me what great choices I’d made and that she was looking forward to me reviewing them once I’d had a chance to try them out.
Millie’s note in my delivery was much more aligned with my expectations as a customer than the emails.
These recommendations continued to come on a weekly basis which is over-the-top for my once-a-year bulk shopping habits so I’ve ended up turning them off. Now, if they slowed down after a purchase to being sent once a month or so, it’d feel much more natural.
All that being said, I’ll be sure to return to Thread. I’d go as far as saying it’s the best shopping experience I’ve had in years. A genuine mix of styles, that were cleverly presented with shoppable lifestyle images that inspire and provide a clear route to success for the "fashion challenged" like myself.
My fresh Threads.