The Problem(s)

The Windows Dev Center initially combined the Windows 8 and Windows Phone app store backends into one single portal. This allowed the publishing of apps and games using the Universal Windows Platform (UWP). As an “App Store” in the 2012-era definition, this meant somewhat simplistic products (when compared with AAA games). When the business decision was made to unify the Xbox and Windows stores, the Windows Dev Center was left lacking key functionality and features required for large games studios. To meet an arbitrary early deadline, the decision was made to layer that functionality into the Windows Dev Center, which was built for much simpler scenarios. What resulted was slightly disastrous, with partners hesitant to onboard and instead sticking with legacy Xbox dev portal (which also won’t be winning any UX awards, but was basically reliable, albeit not self-service friendly).

Some of the key issues with Dev Center included products needing to be submitted and instanced as a singular cohesive “submission,” slowing down iteration, publishing, and updates. The lack of an infrastructure flexible enough for collaboration (up to 1000 individuals can work on a single AAA game) introduced process, delays, and dependencies to an already complicated life cycle. Different personas work on different parts of the shipping game. Microsoft was forcing friction on these users by forcing them to coordinate at otherwise unnecessary points in the development cycle.

All of this was learned through observation, consultant research, forum and support feedback, dozens of interviews with developers of all sizes, as well as first and third-party dev liaisons.   

Hypotheses

With the exceptional amount of research accumulated, I put together dozens of hypotheses - ideas with a measurable result. Realizing that Dev Center’s problems reached far beyond the UI layer, many of my ideas necessitated a re-architecture of the underlying systems. At the core of most of the work was the “flattening” of the aforementioned Submissions, which would save us from our hub-and-spoke navigation nightmare and allow developers to directly navigate to the areas they want to work on. Individual contributors would be able to work independently within their studios and fork configuration at a feature level, rather than product-level. This enables a late-binding publishing system that allows users or teams to work independently up until a first retail publish (often months or years into the process), collectively publish that first public version, then update flexibly.

A number of opportunities open up, including the ability to service a product in retail while iterating on content updates, without the two versions affecting each other.

Alignment

Design-driven projects aren’t especially common at Microsoft, where the vast majority of initiatives are spearheaded by engineering-centric program managers. The result of this approach, and a “Minimum Viable Product” mindset, rarely yields strategic overhauls in favor of shipping something (anything) in a given engineering iteration. Asking for an engineering commitment for a project that could take two years to build was ambitious. After synthesizing my research and mocking up some initial flows and wireframes showing off the new approach (and explaining what outcome driven design is), I presented to the Dev Center engineering leadership. To my pleasant surprise, within weeks they assigned a PM to drive this full time.

In working with him, we not only came up with a fun name (Project Jaguar: fast car = fast iteration) but were also able to stress test my initial thinking and account for the intricacies and edge cases of the games industry.

Microsoft sometimes “ships the org chart,” a phenomenon in which an experience’s engineering seams are very clear to the end user. Through a strong partnership with the Xbox Live PMs and engineers, we were committed to not letting this happened. Xbox Live was the area I was most passionate about when I was focusing on feature-level work, but I was able to hand the bulk of that work to a designer on my team. She and I closely collaborated on some of the most complicated AAA game dev scenarios while delivering designs that scaled for edge cases, without designing around them.

The partnerships we built had an improv sensibility - we applied a “yes, and…” approach to each other’s’ ideas, resulting in some of the most fruitful collaboration in my career.