All games on Roblox are UGC -- built using Studio, Roblox's game engine. As a UX design intern at Roblox this summer, one of my main projects was to research and define proto-personas for Studio, because a major block for the team and product planning has been the lack of agreement from stakeholders on who the target audience should be.

The main objective of my project was to align the visions of all stakeholders, to form a set of proto-personas that everyone supported, as well as to help the Studio team adopt customer-centric thinking.

My game plan for this project would have to answer these questions:

Under the guidance of the player-side UX researcher at Roblox, I planned a 3-part proto-persona workshop, with the key stakeholders of the product, design, and engineering teams of Studio as the participants. We decided on this approach because the workshops would allow stakeholders to share and get feedback on their different visions for Studio’s target audience. Then, based off the stakeholders’ own proto-personas, I would draw a set of proto-personas for everyone to verify and ultimately choose who to support as the primary target audience. 

After this, the next steps would be helping everyone understand how to use these proto-personas in their respective functions.

How would I ensure that this lengthy workshop would be a valuable investment of busy stakeholders’ time? To answer this, I held 1-on-1 meetings with individual stakeholders to understand their current thoughts on Studio’s target audience, the value of defining our customers, and how personas could help their work. This proved effective, as all stakeholders agreed to attend the workshop.
‍

Based off my conversations with stakeholders, I realized that the concept and application of personas was quite new to most of them. As a result, I prepared a short presentation that highlighted how a team at a company might use personas to prioritize features, plan the product roadmap, and segment the market -- essentially, making clear the value of proto-personas and personas to Roblox. Below are a few slides from the deck created.

After introducing personas to them at the beginning of the workshop, the 9 stakeholders were asked to create proto-personas to represent their ideal target audience for Studio. Altogether, they created 18 different personas, modeled after "Jenny," the proto-persona that I envisioned to be the primary target audience for Postmates, as seen below.

Then, each person introduced the “friends” that they had created. To help drive home the idea of being empathetic to our customers, I asked them to present their proto-personas as if they were introducing a friend to another friend and encouraged questions about realisticness and why someone decided to prescribe a certain attribute. Overall, the exercise was actually pretty fun, and the conversations were insightful. In fact, a few days later at a team meeting, one of the engineering directors brought up the necessity of defining Studio’s target audience with all the engineers -- a first!
 

The next part of the workshop focused on placing all the new “friends” onto attribute scales so that they could eventually be clustered. To find these scales, the senior UX researcher and I affinity mapped all the attributes found in the 18 proto-personas before the workshop, and found that these could be grouped into 5 different categories.  

The voting was quite a lengthy process, since stakeholders voted 5 times for each persona. As the host, I had to gauge the energy of the room and know when to encourage discussion, because I wanted to make sure stakeholders’ interest and investment in these personas stayed high. By the end, we all agreed on where these 18 personas fell on these scales in relation to each other. 
‍

The next step for me was to cluster the 18 personas into buckets, based off the similarities in their data points and the unique qualitative attributes the stakeholders had given them. Four distinct groups emerged from the clustering; I chose these categories while keeping in mind what kind of grouping would best inform the Studio team in planning the future of the product. I created proto-personas to represent each grouping.

With the same group of stakeholders, I presented the resulting set of proto-personas, and the team reviewed how realistic they were and their individual qualities. After a few small changes made to a few of the personas, the team was in agreement on the four different kinds of target audiences that could be pursued, and was ready to vote for who they supported. I structured the voting such that each stakeholder had 3 votes to distribute however they liked -- that way, they had to eliminate at least 1 group, and depending on how strongly they felt, they could champion another group.

After voting, I opened up the floor to discuss how people voted, and then allowed people to change their votes. The discussion was insightful and a few stakeholders shifted their votes at the end. At the conclusion of all three sessions, the team had successfully aligned on who to support as their primary target audience, a secondary target audience, and an anti-persona -- a huge step for a team who was unsure about the value of personas a few weeks prior. 

Although my internship ended, I initiated the socializing process by creating sets of portable cards for the stakeholders of Studio. These cards were two-sided; one side was to contain quick, digestible information such as the drawings of each “friend” and words to describe them, and the other contained the rest of the information on their demographics, behaviors, and needs. 
‍
The team also showed enthusiasm in adopting a more user-centered way of thinking. To introduce all functions to utilizing personas, the design team plans on bringing the cards into all new projects to ensure that each project has a target persona. Moreover, the design process will now include a phase that researches who exactly the target users for each project are. 

Previously, the value of narrowing our target audience to specific groups was not understood by the entire team, the concept of personas were a mystery, and key stakeholders did not share the same assumptions and beliefs on who Studio should be built for. This project was a step to becoming more user-centered. By understanding how personas could improve each function’s workflow during the 1-on-1 meetings, I used these insights to help stakeholders align their diverging perspectives, which would ideally then trickle down to the rest of the Studio team.
‍
As an intern, navigating the office to push a new kind of product thinking seemed daunting at first. I knew I needed guidance, so I sought the mentorship of the sole UX researcher at Roblox and other members of the Studio design team to provide me the proper insights into the team to determine the best steps for this project. By the end of my internship here, I’m confident that the socialization process of the proto-personas will have begun the slower, ongoing movement towards a user-centered design process.