Inside Figma's Push to Automate Research Operations, Not Research Judgment

The following customer story is based on our AMA with Figma's research team, featuring Christa Simon, Product & Strategic Research Lead,and Peggy Zhang, AI Insights Program Manager, Research.

For years, some of Figma's most valuable research insights lived where many research teams keep theirs: scattered across reports, buried in FigJam files, and stored in the heads of researchers who have accumulated years of context.

When teams needed to understand what the company already knew about a customer problem, finding the answer wasn't always straightforward. Researchers were often pulled into conversations because they alone remembered where insights lived. Because existing findings could be difficult to surface, new research was sometimes commissioned even when the organization had already collected evidence that could inform the decision.

The challenge was making existing research accessible to everyone across the company. As Figma's products, teams, and customer base expanded, they needed a better way to preserve institutional knowledge and make insights available to the people making decisions. That need eventually led to two parallel investments: a dedicated AI Insights function and a recruiting infrastructure built to support research at scale.

Managing Participants at Scale

As Figma's research program grew, so did the complexity of managing participant relationships, recruiting activities, and study coordination. Spreadsheets and ad hoc processes weren't cutting it.

Figma turned to Rally to build a centralized research CRM for participant management. Researchers could access participant history, coordinate recruiting efforts, and run studies without the operational overhead that had been slowing them down.

That foundation mattered more as the research workflow evolved. When researchers began building automations that connected multiple systems, Rally served as the source of participant data those workflows depended on.

Creating More Time for Research That Matters

When Peggy Zhang joined Figma's research organization as AI Insights Program Manager, a title and function that didn't exist five years ago, her mandate wasn't to replace researchers with AI. The focus was far more practical: automate the repetitive work surrounding research so researchers could spend more time on synthesis, recommendations, and collaboration.

"We really wanted to leverage AI to automate the tasks that are not really research-thinking related," Peggy explained. "[Researchers] are able to focus on those things that matter."

That philosophy shaped every decision that followed. Rather than searching for isolated automation opportunities, Figma began by investing in the systems needed to support automation at scale.

"We spend a lot of time thinking about the infrastructure to make that happen," Peggy said. "We want to empower all Figmates to build their own tools, but in order to get there, we need to set up the infrastructure for that."

The goal wasn't for a small team to build every workflow, but to create an environment where researchers could build, share, and improve workflows themselves.

Building the Systems Before the Automations

Many organizations begin their AI journey by experimenting with tools and workflows. Figma took a different approach. Before introducing large-scale automation, the team focused on organizing information, creating reliable sources of context, and establishing places where workflows could be shared.

Peggy describes the effort as building a "Design and Research OS," a shared environment where researchers and designers can publish AI skills they've created and make them available for others across the organization. A workflow built by one person doesn't remain confined to one project; it becomes something colleagues can install, adapt, and improve.

That model has helped Figma move beyond individual productivity gains and toward shared operational improvements. The infrastructure makes the automation possible.

The Workflow One Researcher Built for the Entire Team

One of the most compelling examples came from a workflow created by Neha Kodi, a researcher at Figma. Using Rally MCP, she built an automated AI skill designed to eliminate several recurring administrative tasks that happened around every research session.

Before a session begins, the workflow automatically sends a briefing via Slack to the project team, including participant metadata, screener responses, and observer links. Researchers no longer need to gather and distribute that information manually. After the session concludes, the workflow pulls together recordings and the associated Slack discussion thread to generate a synthesis summary that helps the team quickly review what happened.

What happened next is what changes Neha's story from one of convenience to one of transformation. She didn't keep the workflow for herself. She published it to Figma's shared repository so other researchers could install and use it in their own projects. A solution built for one researcher became a resource for the entire team, standardizing best practices and multiplying efficiency gains across the org.

That pattern is increasingly common inside Figma: researchers build workflows, share them, and create leverage for colleagues who may never need to solve the same problem from scratch.

Democratizing Research Beyond the Research Team

As access to insights improved, the conversations around research at Figma began to change. According to Peggy, stakeholders previously spent much of their time trying to locate relevant reports and findings.

"Before all the AI workflows, people were asking, 'How do I find a report related to a topic I care about?' Now people are asking, 'I see a couple of reports related to this product area I'm working on. What is the most important insight surfaced from that?'"

That shift has implications well beyond the research team. Product managers, designers, engineers, and other stakeholders can access existing knowledge more easily and engage with research earlier in the decision-making process. The conversations move from locating information to understanding it.

What AI Changed, and What It Didn't

As more operational work became automated, researchers found themselves spending more time in areas where human judgment remains essential: methodology, interpretation, stakeholder alignment, and building trust in findings.

Christa believes those responsibilities have only become more important.

"The job of a researcher is ~80% relationship building," Christa said. "I don't think that changes."

"AI doesn't replace research judgment, it raises the bar on it," Peggy added.

Researchers still decide which questions matter, how studies should be designed, and how findings should influence product decisions. Automation accelerates parts of the process, and it also increases expectations around the quality and clarity of recommendations.

What's Next for Figma

Figma's next area of exploration focuses on one of the most persistent challenges in research operations: recruiting highly specific participant audiences. Peggy pointed to Rally MCP's capabilities as an opportunity to connect participant data in Rally with information stored elsewhere across the business.

The vision: a researcher identifies the audience they need, systems pull together relevant data from multiple sources, and qualified participants can be identified and recruited programmatically, with far less manual effort. For niche recruitment efforts that previously required extensive coordination, the impact could be significant.

The same philosophy that shaped Figma's broader AI strategy continues to guide what's next: build the systems that allow information to move efficiently across the organization, then create workflows that help people spend less time managing research and more time learning from it.

How Rally Can Help

Figma's approach is replicable, and you don't have to be at Figma's scale to start. Rally gives research teams the participant management infrastructure they need to recruit, coordinate, and run studies without the operational overhead. Rally MCP makes it possible to connect that participant data to the broader workflows your team is building.

Book a personalized Rally demo today and see where it can take you.