Improving the discoverability and inclusivity of a large product range.

I was responsible for validating, designing and supporting delivery of a new product navigation in order to launch new products and reach new audiences.

Product IA & Navigation
iTunes App StoreGoogle Play StoreView Website

Made one at a time, printed just for you… but hard to find.

The stickers, t-shirts, posters, phone cases and 60+ other products sold on the Redbubble marketplace are printed on demand, which means that they never go out of stock. These products had been selling for years without much change to the system underneath (aka, it was all hardcoded). And the 2019 company strategy was to kick off a huge number of product launches in order to acquire new customers.

While I was working in the Customer Acquisition team, I raised the question, "as our product launches scale, what parts of our experience don't?". A quick audit of the main shopping journey confirmed my fears - we had a lot of inconsistencies in the way we name, structure and categorise products. You could even see 3 different names for the same product across the experience. This wouldn't just be an engineering concern, it was a big usability issue that would only get bigger and harder to manage.

With a limited scope of 6 weeks, myself and a rockstar researcher (Jo Lumsden) set out to create & validate a new product hierarchy, product navigation tree and product name recommendations, with consideration for cultural differences in our European markets.

The goal was to improve discoverability of our products, which would then give us a boost in SEO ranking, and trickle through to improve conversion as well, as more people were able to find a product to suit their needs.

I don't believe in edge cases.

I had a personal goal to make our clothing navigation less gender binary. Who were we excluding by defaulting to Mens & Womens? Our customer base is global and diverse. It also limited the number of categories we could have on our horizontal product nav on desktop and made it harder for us to categorise unisex or non-gendered products - we were forced to give products, like scarves, a gender. It was a lose/lose situation.

We made sure that in our research recruitment there were people who identified as neither male or female, but we also found that cis-gendered participants were questioning the need to navigate by gender as well.

"It was frustrating having to divide the t-shirts based on gender. Ultimately I decided men / women because that’s what it’s like on most websites — But anyone can wear whatever they want. I would want to see men, women, gender non-conforming models to see how they look for all potential buyers."

Card sorting participant.

Making it easier for customers makes it better for the Googlebots too.

Speaking to internal stakeholders, from Commercial & Marketing to the Design System team, we learnt that displaying the same product in more places, potentially with different names and categorizations, could help drive more traffic to the site.

I then started to gather insights from Google search console and keyword planner. This informed some of the names of our product departments and whether we needed to merge or split product categories in order to be competitive in our rankings in Google search.

To get a customer's perspective, we conducted card sorting exercises, where users were asked to sort our products into categories within only a name and basic description.

This opened up new opportunities to branch into potential new product categories, and gave us direction and inspiration for grouping products.

We also learnt that context is everything. Customers change the way they browse and shop based on who, what or why they are shopping for online.  Contexts that stood out to us were shopping for kids or personalised gifts. This navigation needed to handle all these cases.

Tree testing allowed us to test assumptions in our draft product tree. We learnt so much from the first round that we iterated and conducted another. Ultimately creating a navigation that easy and intuitive for customers, and also helped us rank in organic search.

When a new product is launched, how discoverable is it?

Redesigning the header to showcase the brand, and providing users with a nested navigation allows us to scale the product groupings as much as needed. I built a mobile prototype and tested the usability of the validated product IA in context. We learnt that imagery within the menu helped customers to quickly scan the list for the category they wanted. And presenting the navigation as a side drawer gave them context of where they were on the site which made them feel confident to continue shopping.

With a design principle of "mobile first", we could have stopped there. But I believed that the desktop experience had the opportunity to be different, and with our shiny new global nav API (that I helped prioritise with engineering leads), we could present our product groupings in many different ways, and also easily change or add new navigation items.

A mega menu allows Redbubble to also showcase popular topics and promote recently launched products. Customers found it easier to scan and it invited them to explore other pathways they wouldn't have normally found.

Impact & results

With our new product IA and navigation, we succeeded in helping visitors connect with Redbubble and find what they are looking for. We saw a 3% increase in conversion, and an 8% drop in exit rates.

Displaying the "zipper pouch" product with an additional name and pathway, "pencil case", was highly impactful. It resulted in 4% growth in organic traffic, proving that context matters when customers are searching and our navigation should respond to these needs in order for us to be discovered on Google.

Customers noticed the gender inclusive change within days of release. My team and I welled with pride when we found this tweet and the ones that followed it. Redbubble is a strong supporter of LGBTQIA+ rights, so we all appreciated the impact our work had on people who were non-binary or gender fluid. It also didn't negatively impact conversion which many people raised as a concern.

The knowledge learnt from this project has sustained my work for the past 12+ months - it's still helping to inform product launches and still scales as we launch more products. In order to achieve this cross-team collaboration, I built relationships with non-engineering teams that meant future projects were even more rewarding, efficient and collaborative!

Want to chat more about this project and other things I’ve worked on?

More of my work

Yackandandah Folk Festival

Designing a festival app for a small country town with a big heart.
View Project


Helping property planners find planning schemes, overlays and permits.
View Project