Redbubble observed a considerable drop in organic traffic due to a Google Search algorithm change. The biggest impact was felt in t-shirt sales, which was compounded by a flood of new competitors in the market. A new team was founded to investigate what was going on, improve landing experiences for t-shirts and create a user-centred SEO strategy for the company.
Internal links give Google an idea of the structure of your website. They can establish a hierarchy within your site, allowing you to give the most important pages more link juice than other, less valuable, pages.
But the current state of Redbubble’s internal links was that we were telling Google that super niche pages were the top of our funnel, which wasn't great, and was resulting in a large number of exits. Ideally, we wanted our link hierarchy to start with top level product categories that all other t-shirt related pages linked to. But these pages didn't exist yet. That's where I came in!
What is important to users when first landing on Redbubble?
What are they most likely to do next?
I spoke to people who had never visited Redbubble before, starting them on a Google search result page and getting them to click through to Redbubble, I collected observations of their first impressions.
I turned these insights into a problem statement, that I then used as a prompt for a co-designing workshop with the whole team to bring everyone on the entire design journey.
To have maximum impact, we felt that creating a new landing page for tshirts would be a great place to start so we could start to measure improvements to SEO, knowing that it may take a while to see a shift.
In my exploration, I focused a lot on what could be reduced, breaking down the relevant components by their parts to understand how effective they were for landing experiences. Some of the things I explored were:
Purposeful hero banners
The hero banner works to communicate to users what they're going to see, and whether it is relevant to them. It also is a great chance for us to inject a bit more brand personality into the experience through photography.
Our marketing photography set was quite limited and was created for a specific context. So instead, I explored using influencer imagery in prototypes to test whether they resonated with customers and our brand personality. Consistently, participants said that the imagery conveyed our brand as diverse and modern.
I also explored how these banners might work for more of the niche keywords that we rank for. These hero banners also needed to have some kind of dynamic content. I explored extracting the first product in the results and featuring it as a hero, playing around with colours from the design or even blurring it, and also using the computer generated imagery at our disposal.
Based on the context of where the user lands, the different modules of the page could be switched out depending on the users needs and likely pathway at that point in their shopping journey.
I tested the new landing page experiences with users through unmoderated user testing, so we could collect perspectives from our diverse, global audience. I got the team involved by assigning everyone participant videos to watch, and then we did analysis together, even remotely though Miro and Trello.
For an agile team, waiting 2 months to see results was pretty stressful, especially given the amount of eyes that were on our project internally. But because we had validated early and often, we were confident we would see improvement in the user experience that would then translate into SEO wins.
And it began to pay off. 2 months later, we began to rank #1 in the US for some top t-shirt related queries. On average, we now show up on the first page of results for t-shirts, meaning we also saw an increase in new users visiting the site. We saw a huge reduction in exits as engagement on site increased by 45%.
Because of the success of the new Landing Page experience, I pitched a strategy that we should roll out the new experience to other departments, and empower non-engineers to build these pages and pathways through a CMS. This would mean that future changes wouldn’t require any engineering effort.
I built guidelines on usage of components that were then added to the Design System to help commercial and marketing to build the pages themselves. Our team built modules and templates that were editable in Contentful, and then similar landing pages were rolled out to all the other product categories. These landing pages are also now used for marketing campaigns, licensed fan art pages & gift guides across the Redbubble experience.
Giving users different ways to get started when they first land on Redbubble when shopping for a t-shirt was a successful strategy that ultimately resulted in improved SEO, and improved user experience.
