Our old marketplace funnel performed rather poorly in the number of offers and lists made. This has a direct impact on our sales as the more offers and lists that are made, the more likely a transaction would occur. As we spent a considerable amount of investment in advertising and marketing to bring traffic to our marketplace, having a low conversion ratio would be quite wasteful. Therefore we made the decision to optimize the marketplace flow.
We started by looking into the site analytics to identify dropoffs and potential problems our users are facing. To start with, we tracked the funnel and map it to the user’s actual journey across making a list, offer or purchase.
Through this finding, we can see a significant volume of visitors dropoff at the start of the journey.
We also see a huge amount of support tickets related to their experience with the marketplace flow.
We conducted usability tests to directly observe our user’s behavior throughout the process. We reached out to Sneakerheads who are not familiar with our platform but with a basic idea of a sneaker marketplace platform. Since we already have identified several problems with the funnel, we have a clear idea of the goal of the usability test as well as the tasks to evaluate. We also complimented the usability tests with several interviews with sneaker resellers to better understand their motivations, goals, and pain points when making a sale.
Overall we found that the initial flow barely solved any real user problems. And when they do such as making a purchase, it is done in a way that doesn’t match with the user’s intuition.
were confused with the steps to purchase as the interface did not connect to their expectations coming from other eCommerce platforms or even their idea of a marketplace. Their intuition is to find their ideal size then add to a cart. Being blocked from seeing the sizes and being asked to make a decision to make an offer or buy now from a dialog seems bizarre as they haven’t even selected their size yet.
find that making a list or finding offers were too hidden and require a lot of U-turns.
We prioritized user needs that would align most with our business goals and we kept promising but not important ones on the backlog. Then, we began ideating on potential solutions to address them. At this stage, we were mindful of the usability issues highlighted with testing and we seek to avoid making the same mistakes. We used a few days of design thinking to explore many potential solutions before narrowing them down to those that work the best. Our colleagues from the different departments were invited to these workshops so that we could get diverse and sometimes contrary ideas.
Next, we took the top solutions and sketched them into rough wireframes to test with our colleagues outside of the design team. We went through a couple of quick rounds of this before testing with users.
We made several low-fidelity prototypes to test with our users. As we’ve already gone through several internal rounds, we decided to conduct 1 or 2 rounds of usability tests with our users. The first test was to validate the solution with our users while the second was to iron out any usability issues the team may have overlooked.
The first test was a success and our users gave us very positive feedback on their experience with the new flow.
The second test was done on a mid-fidelity prototype, this time we wanted to simulate real usage as best as possible so we made an interactive prototype on Figma for users to try on their devices.
The mid-fidelity prototype paved the way for the implementation. From testing, we determined that there were no breaking usability issues and users generally loved the change. We moved into the final UI and I worked together with our UI designer to incorporate the new branding into this project. We worked on a per-component basis as the layout was already determined in the mid-fidelity design stage. This was also done in preparation for a design system implementation in the future.
As this was the first large-scale UX project within Novelship, we made a lot of trial and error along the way. There was also the lack of manpower while given a very limited time to solve a complex issue so we had to innovate and make certain shortcuts. We had no real way to talk to users so we had to build a system to reach out to users and worked with the Customer Support team to funnel in tickets that highlight usability concerns. Internally, we had to manage stakeholder expectations and we tried as best to include them in the conversations we had.
To understand the impact of the change we made on the marketplace funnel, we implemented additional tracking and surveys to collect more data on how our users are utilizing this change.
