Evaluative research of a financial wellbeing platform for Meta employees

The situation

One of my responsibilities at Round Feather was to grow our evaluative research capabilities. I got the opportunity to do so when Meta approached us to explore and solve perceived employee benefits issues. Our first round of research was explorative and led us to better understand why employees would not take advantage of all company benefits and perks available to them.

Our research efforts revealed the following key insights:

• Benefit systems were spread across disconnected providers;

• Employees experienced knowledge gaps and a lack of personalized guidance;

• Confusing terminology and limited awareness were barriers to engagement with the platforms;

• Employees were dealing with the emotional burden of financial stress in their high-performance environment;

Company reports showed that these issues led to an estimated $300M in productivity loss.

Round Feather's emotion-driven research showed the need for reassuring employees, making them feel safe and confident about their financial situation, goals and decisions, and guiding them to these benefit platforms they were unaware about. The solution turned into designing a financial well being app (i.e., Paven) that needed to continuously be tested against all proposed features and capabilities.

My role, team and responsibilities

As a UX Research Lead, I was responsible for collaborating with the design and development team, and driving the evaluative research activities for Paven accordingly. I teamed up with a Design Researcher to ensure cross-validation and quality of findings, and timely deliverables to the team. I divided up the story into three milestones below. Altogether, I planned, executed and shared insights of 8 heuristics evaluations and 25+ user testing sessions.

The story of Paven

Paven as a concept: gathering feedback in the field

I joined Round Feather in 2019 and got baptized by fire on the first week. I was sent to Menlo Park with some of my new colleagues to introduce Paven as a proof of concept to Meta's early adopters. The initiative lasted a few days and was packed with meetings with Meta employees. Our goal was to understand how the proof of concept and its potential value would be perceived by employees.

I was brought up to speed the week before about the preliminary, exploratory research that was done by the team. It gave me a good amount of information and background so that I could design the first evaluative research plan for this field trip. We designed three research-informed stories featuring three different people with different backgrounds and life moments. Each story went a different path, aiming to showcase key features of the proof of concept.

This proof of concept was deterministic: the experience flows and interaction levels were limited, so I decided that the sessions should be run as guided walkthroughs as opposed to task-based. The sessions were mostly scripted to ensure consistency across all of them. Each feature walkthrough was paired with moments of silence or specific prompts to enable flexible feedback at the right time.

Great, meaningful feedback was received from more than a dozen employees. I sensed a lot of excitation and desire to contribute to the creation and growth of the concept, which was a big win for us. I ran some analysis when I got back home and presented the results to Meta's benefits leadership.

Paven 2.0: ensuring key features will be useful

Once our proof of concept was perceived as meaningful by Meta early adopters and leadership, the objective was to develop a first version that would align with user needs. Design, testing and development teams joined efforts to bring Paven to life in the most human-centered way possible: prioritization of features against research-informed needs, rapid prototyping of prioritized features and user testing with Meta employees, translation of research insights by the testing team to the design and development team to ensure there was no loss in translation. 

During this initiative, I led the testing of 15+ features and associated task flows with Meta employees over the course of 3 months. Each week was focusing on one specific feature. One testing week included an alignment meeting regarding the feature to be tested, the design of the research plan, the testing sessions, the analysis of the testing sessions and the sharing of the insights to the team.

This cadence allowed me to stay focused on the research goals, success metrics and whether the prioritized features to test would reach these metrics or not.

Paven 3.0: understanding low levels of engagement

After we launched Paven to Meta employees, the development team was carefully monitoring the its efficiency. While it has seen growth in adoption, it has faced retention challenges. Analytics suggested that some features were less engaged with. Indeed, pushed notifications and its homepage were the main engagement drivers.

The development team documented these insights and put together a plan of action to improve the application. I jumped in at that time to help them figure out why engagement levels were low. This phase included 7 initiatives in which I conducted secondary research, heuristics evaluations and user-facing testing to gather relevant insights that were shared with the development team.

The overall impact of Paven so far

• 14,176 users onboarded since launch

• 83% of users set financial goals in-app

• 7,554 benefits connections initiated via the platform

• 44% of tax advising appointments and 58% of financial coaching engagements happened via Paven

• Improved employee satisfaction and measurable ROI on benefits programs

The platform reconnected people with resources they already had access to—just in a more human, contextualized, and actionable way.