In late 2023, our app and brand were expanding quickly, yet our website lagged, misaligned with our current branding. It was an unstable mix of eye-catching banners and loud callouts, yet failed to clarify basic queries about CGMs and our product, leading to increased support tickets. The challenge: could we address these issues and boost conversion by 10%?

Goal & Objectives 

1. Enhance Brand Alignment: Redesign the homepage to reflect the premium nature of the product. 

2. Clarify Offerings: Be clear about the differences and benefits of CGM subscriptions and Levels memberships. 

3. Improve Narrative: Personalize to resonate with target audience and illustrate the problems Levels solves. 

4. Increase Conversion Rates: A more intuitive and informative story should improve higher engagement and member sign-ups 

5. Reduce Support Tickets: Address common questions and confusions through clear, accessible content to lower support tickets.

Hypothesis

Redesigning the website to align with current branding and providing clearer information about CGMs and Levels will reduce support team strain and increase signup conversion rates by 10%.

Design Process

We had limited resources for this project, so I suggested designing and implementing the test with Framer to avoid the need for additional front-end help. This approach would enable us to design and implement with minimal or no coding, allowing us to quickly test our hypothesis and determine whether further investigation was warranted.

In addition to improving the narrative to better reach our target audience, we needed to address our brand inconsistencies. To do this, we designed the new website test using components that could be reused and improved if they proved successful in the A/B test. Component-based UIs provide good constraints for any further internal or external website work, particularly for those not as familiar with the design system.

Furthermore, I designed two new sections to help our support team save time. The first focused on CGMs, explaining what they do and how they can benefit customers. The second detailed the Levels membership and how it integrates with the CGM.

Challenges

We faced setbacks in our A/B testing phase, having to relaunch the test twice due to critical bugs that invalidated the results. These issues caused nearly two-week delays in the overall project timeline. By the time we launched our third candidate, the results provided inconsistent data which ultimately diminished our confidence in the final test data.

Outcome

Our hypothesis was ultimately invalidated, as the new website converted 12.5% worse than the control. It's worth noting that within the signup flow, the group tested with our Framer prototype showed a slightly better conversion rate. Despite this, the overall performance of the new website did not meet our expectations, leading us to conclude that our initial hypothesis was incorrect.

Learnings & Takeaways

1. Efficiency with Framer: Framer allows for quick iterative development, reducing the need for extensive engineering resources.

2. Limited Impact of Brand Cohesion: Brand cohesion showed minimal direct effect on increasing signup conversions.

3. Clarifying test variables: As our final test invalidated the hypothesis, data revealed that altering the main CTA's copy and message, despite aiming for clarity and focus, might have reduced its effectiveness. The original banner and its message may have been more effective in capturing visitor attention and driving conversions.

4. Test more, sooner: While we did conduct stakeholder testing before launch, involving a larger team segment for user testing could have identified critical bugs sooner. This would have minimized the need for multiple test relaunches and allowed for earlier result validation.