Figma

The challenge: Pioneering at-scale product measurement

Figma serves over 540,000 customers and needs a world-class data infrastructure. When the Research team collected survey data, they had to manually export CSVs and thread information together with the Data Science team’s help. The process was expensive, time-consuming, and couldn't scale.

AI Insights Researcher, Rie McGwier, joined with an ambitious mission: build NPS and Product Health programs that would turn millions of survey responses into insights the entire company could use. This required live dashboards combining qualitative and quantitative data that non-technical stakeholders could explore. The challenge? Rie was one person, self-taught in SQL and Python, building enterprise-grade infrastructure that would normally require months of learning or drastically reduced scope.

The solution: Building world-class data infrastructure through AI-powered iteration

Rie used Hex's Notebook Agent to compress the learning curve from weeks to hours, focusing on strategy instead of syntax. Figma's survey data flowed from Sprig into Snowflake, where it was triangulated with behavioral and demographic data, then transformed in dbt Labs. Hex became the experiential layer, turning raw and unstructured data into insights teams could access.

"Without Hex’s experiential layer, these programs couldn't take flight — no matter how good the data was." - Rie McGwier, AI Insights Researcher

Rie structured the program in four layers:

• Signal - integrating raw data into Snowflake

• Meaning - creating context in dbt Labs and Hex

• Experience - building accessible insights through Hex dashboards

• Extension - pushing updates to Slack

Enriching data: Building the foundation for automation

Over time, Figma's survey data evolved from simple fields like signup date to complex calculated metrics such as users’ age when feedback was given. Without standardization, teams risked calculating metrics differently, creating conflicting versions of truth. Rie used the Notebook Agent to build reusable components with predefined logic that ensured consistency. This created governed scaffolding that multiple teams could confidently build upon.

Building an insights ecosystem that scales beyond one person

Senior Researcher, Molly Jane Nicholas, used the Notebook Agent to accelerate report development while maintaining her analytical judgment. The agent handled the technical mechanics she disliked, such as building charts and writing queries, while she focused on insights and strategy. Molly even created her own reusable component that prevents surveying restricted domains, which other researchers and data scientists now use. Despite not writing SQL, she can now launch surveys, analyze feedback, and build dashboards in minutes instead of days. Her Hex reports have become core to the company's business reviews.

Dashboards and insights delivered where teams work

With the new data infrastructure, product Health dashboards now refresh hourly, giving leadership real-time visibility across Figma's portfolio. Product-specific dashboards follow a modular design, guiding users from high-level summaries into deeper explorations of open-ended feedback.

"Hex's ability to develop a notebook view and place visuals in a dashboard is what made the productization of our insights possible," Rie explained.

"The dashboards have become an organic part of how teams are evaluating their products," Molly said. When teams identify problems on the dashboards, they move into FigJam for ideation, then Figma Design to prototype solutions. The cycle repeats: measure in Hex, understand insights, design in Figma, validate, build, measure again.

The extension layer proactively delivers insights through weekly Slack summaries. Information flows to teams where they work, sparking conversations that lead to product improvements.

The Impact: Best-in-class customer survey reporting

Figma's continuous measurement programs have become the backbone for product insights across the organization. Rie's infrastructure increased both access to information and trust in data sources. Teams now have governed, real-time insights at their fingertips instead of waiting days for analysis. The impact is visible everywhere: dashboards appear at leadership offsites, product teams use them for daily decisions, and researchers build on the foundation to launch new measurement initiatives in minutes. What one self-taught researcher built with AI assistance now serves thousands of Figmates.

With Hex’s AI, the Research team at Figma:

• Reduced the learning curve for Python and SQL from weeks to hours

• Enabled the team to track 2M+ data points across NPS and product health programs

• Enabled researchers to launch surveys and build reporting in 20 minutes instead of days

• Generated automated insights in hours instead of days

• Built a reusable component that can feed 10+ dashboards

The roadmap ahead is equally ambitious. With Hex, the Figma Research team plans to enable self-service exploration through Hex’s Threads, letting non-technical stakeholders ask their own questions, while automating the remaining friction points where manual work persists. Rie and Molly plan to use the same iterative approach that built the foundation of their data structure to mature their insights over time.