Thread

The service that recommends clothes in your style, size and budget.

  • Product Design
  • User Research
  • Data Analysis

Context

Thread is a service that pairs human stylists and AI to recommend clothes for men in their style, size and budget. I joined Thread as a second product designer and worked on the Growth team, where my primary role was to understand what turned users off of online shopping. Our primary audience was men within the 30 - 45 age range, of who most were engaging in online shopping for the first time. There was substantial untapped potential in crafting an experience for these men who were increasingly moving away from the high street shopping experience due to time constraints and lack of personalised knowledge on what to wear.

The serial comparer

We uncovered one of our most successful customer solutions through a mixture of phone calls, in-person product testing, and user recordings. As we engaged in these different mediums of user testing, we noticed a segment of customers I termed as 'Comparers'.

The typical behaviour of this customer would be as such:

First, they would search for an everyday item online such as 'Light Blue Jeans'. After which, they'd find one of our products in the Display ads carousel of products.

They'd then click into the product, have a brief look and then click back to Google to compare other products. We noticed a cognitive overload develop as customers moved between experiences, frequently re-closing cookie banners and re-finding return policies.

Having one of the largest available brands and clothing range online, it made sense that users would prefer to compare in a singular experience, so we mimicked the google shop experience on our website.

When the user clicked into a product on our website, we would present their original item choice in and amongst other similar items, some of which were cheaper, more expensive or discounted.

In one of our successful variants, the original product is subtly highlighted with a dark strip. In another, we took the customer actions on their behalf by showing them the collection of products for a second and then opening up the product modal, giving the user an understanding that closing the modal will take them back into compare mode.

This experiment successfully beat our control experience on various metrics, the most obvious being a significantly reduced bounce rate and increased time on our website. But, most importantly, it increased ad conversion by 23.7% and was the most successful growth experiment by the time of my departure from the Thread team.

... I'm still pulling all the pieces together to complete this set of case studies. Bare with me!