Designing a platform from zero

The Wayfair carrier platform was more of an engineering entity than a user-friendly platform. Users managed carrier data, such as rates or invoices, through either extremely outdated tools that were not owned by anyone or by creating Jira tickets. Most attempts at optimising the way we used or interacted with carriers were unsuccessful because the engineering team did not have the bandwidth to work on it.

To gain a comprehensive understanding of the problems our users faced in their daily work and their needs, I initiated a baseline research project.

In collaboration with stakeholders, product and engineering managers, I explored the drivers of change that would impact the logistics industry and Wayfair's delivery services in the coming years. At that time, we were witnessing rapidly rising logistics costs due to severe capacity constraints caused by the COVID-19 pandemic. The increased profits in the industry were starting to be invested in better technology and connectivity. Similarly, Wayfair was also investing in logistics and fulfilment, with plans to offer it as a standalone service.

The concern amongst the group was that our current platform was not capable of scaling up to meet these challenges.

In collaboration with the users, we mapped out the ecosystem in which they operated and discussed the challenges they faced. Shipping being a critical aspect of e-commerce, they were often placed in high-pressure situations, trying to resolve issues before they affected too many customers. They also felt a significant amount of pressure from higher management to improve and optimise, which was challenging because they didn't feel they had much control over their domain.

Additionally, there appeared to be a degree of isolation from the rest of the business, with users seemingly more connected to their supplier/carrier side contacts and colleagues from their immediate team.

We mapped the experiences of several users who interacted with the carrier platform as part of their job and found that:

• Most users were building a separate data ecosystem outside of the platform and only informed us of changes at the last minute.

• They struggled to get changes prioritised if they were deemed to be low-impact by the engineering team.

• They had little visibility into what data existed within the platform.

• They spent a significant amount of time documenting the problem rather than fixing it, such as investigating to make a case for prioritisation to the engineering team.

Interviews with users across all levels of operations revealed that most felt they were only working on 'low hanging fruit' and couldn't drive meaningful, high-value change because they were never given enough time. The relationship with the carrier platform and technology, in general, was poor due to years of promises and under-delivery.

This had significant implications for the project in that we would need to rebuild trust while also structuring our design response in a way that took Wayfair's business culture into account, such as short turnaround times and shifting priorities.

My product managers and key stakeholders identified what they believed to be the most concrete and important drivers of change, and together we mapped out the opportunity. We believed that providing end-to-end self-service of carrier data to operations would enable them to respond to known issues and tasks more quickly, freeing up engineering to tackle as yet unknown and less defined problems.

If we could build the user-facing services quickly, we would mitigate the risk of the project falling out of 'fashion' and losing priority.

In a co-creation session that lasted a couple of hours, I led a group of key users, product managers, and engineers to understand what the user-facing structure of the platform should look like and to generate early ideas for features. We agreed on a simple side navigation that roughly aligned with the way operations worked with carriers, such as finance teams and performance management teams. We would then build services to support the outcomes those teams required.

Additionally, we identified other features like better user account management and notifications of issues or reminders.

During our research, we spoke with colleagues who had worked on similar projects at other companies and learned that their relationships with carriers improved significantly when they provided access to data, such as performance data. By constantly sharing this information, they found that their weekly review meetings were more productive as everyone was better prepared.

We confirmed that carriers would be open to exploring such a change in the relationship during our interviews with them. Therefore, we put together a proposal to include a carrier-facing portal on the platform.

Our directors said…

No to the carrier facing portal… but you can try something cheap with the internal version

I proposed and received buy-in from one of my engineering teams to build a sidebar for navigation of the platform. I then began designing the core service screens using the Wayfair design system while simultaneously building a pattern library for the carrier platform. Our plan was to iterate on the designs and design system as we built more services.