Enter Password
Goldman Sachs
You earn influence by changing how an organisation listens
Scaling Insight-to-Action Across a 13,000+ Engineer Organisation
Case Study - Process
2020-2023
How I built a scalable insight system using Qualtrics dashboards that connected 130+ engineering leaders directly with anonymised feedback, reshaping how the firm made decisions
Transforming DevX insights from a research bottleneck into a firmwide decision-making system: enabling direct, ethical access to high-volume qual and quant data, structured engagement from leadership, and shared accountability for improving developer experience across the firm.
MY ROLE
Executive Director / Global Head of User Research, Core Engineering
SHOWCASING
INSIGHT-TO-ACTION · DEVELOPER EXPERIENCE · DASHBOARDS · QUAL/QUANT ENTERPRISE SCALING · TECHNICAL INNOVATION · ENGINEERING CULTURE · ETHICAL DATA GOVERNANCE · QUALTRICS
Why This Work Mattered
The Developer Experience (DevX) research programme generated one of Goldman Sachs Engineering’s richest employee datasets, particularly through the annual DevX survey. Scale of collected data would increase year-on-year, but this data lived solely within the research team.
The most transformational thing I did at Goldman Sachs wasn’t running research.
It was re-engineering how 13,000+ engineers, product managers, and senior platform leaders shared, accessed, interpreted, and acted on evidence.
The issue wasn’t the data. It was how the organisation accessed it.
In a deeply engineering-led organisation, you earn influence by changing how people listen.
Background context
The annual DevX survey generated a vast amount of rich, candid customer feedback, which was exactly what we wanted. But the scale created new problems.
Annual DevX survey:
6400 engineers responded (>50% participation)
35 questions - 27 quant, 8 open-ended qual
13,000+ open-ended responses each year
Engineering leaders relied on assumptions, outdated narratives, and local anecdotes to make strategic decisions, while researchers were overwhelmed by the volume of qualitative feedback.
The organisation needed a way to transform this research into strategic clarity, and fast.
If we wanted leaders to make evidence-informed decisions, they needed to be able to see, explore, and trust the data themselves, without compromising anonymity or feeling overwhelmed.
Challenges & Constraints
Researchers couldn’t possibly manually analyse every correlation, and platform teams couldn’t get close to their data without us.
Insight bottlenecks
Manual analysis couldn’t keep up with tens of thousands of open-ended comments
Roadmaps were set within weeks of the survey closing
Visibility & trust challenges
Leaders had no access to raw sentiment
Insights were filtered, slow, and mediated solely through researchers
Governance challenges
High need for anonymity to protect engineers
Strict internal compliance around data access
Structural challenges
23 platform teams with differing needs
Multiple divisions and engineering initiatives
Strong expectation of evidence, but no mechanism for distributing it
The Turning Point
It became clear during the first DevX cycle that researchers alone could never scale analysis at the pace engineering required.
To fix this, we had to invert the model: instead of leadership relying on researchers to interpret reality, we needed to give leaders safe, direct, contextualised access to the evidence itself.
Dashboards became the mechanism - not to display data, but to change the way the organisation listened.
What I Did
1. Built the researcher “mothership” dashboard
I created comprehensive dashboards for researchers that combined quantitative charts, correlation modelling, sentiment analysis, and open-ended text exploration.
This reduced analysis time from weeks to hours and allowed the team to explore hypotheses rapidly while maintaining methodological integrity.
2. Brought leaders directly to their own feedback
I created a template to tailor dashboards for each of the 23 platform teams, filtered to show only feedback relevant to their product area, with carefully controlled permissioning.
For the first time, teams could read raw feedback safely, easily, and at scale.
3. Permissioning, filtering, and data governance
I designed a framework that allowed leaders to access detailed feedback without risking anonymity. This was done by stripping identifying data, using curated keyword filters, and assigning strict dashboard-level permissions.
This became the ethical backbone of the system.
4. Expanded insight access beyond platform teams
During the second year, I piloted dashboards for business units and divisional leaders, giving them visibility into how their engineers were experiencing tools, processes, and changes.
By year three, I had created a version that could be deployed firmwide in only minutes.
5. Increased engineering engagement
I built real-time dashboards tracking survey participation across BUs, reducing hundreds of inbound requests to the research team and significantly increasing leadership and divisional engagement through visibility and gamification.
6. Reduced dependency on researchers
Every dashboard had a corresponding explainer page detailing: how to navigate, how to interpret metrics, how to filter comments, and what keywords were used.
This enabled true self-service at scale.
Mockup of what a BU dashboard looked like in Qualtrics
How things shifted
The dashboard ecosystem reframed how insights flowed across engineering.
We moved from a single interpretive bottleneck (the research team) to a distributed insight model where platform owners, divisional heads, and senior leaders could all access the evidence they needed — safely, consistently, and at scale.
Roadmaps, OKRs, and prioritisation shifted from assumption-driven to evidence-led.
Impact of the work
Cut analysis time dramatically through automated correlation charts
Enabled researchers to deepen insight instead of manually wrangling data
Increased analytical depth and quantitative fluency
More time for strategic research
More efficient annual survey cycles
Elevated the team’s strategic presence
For Stakeholders
130+ leaders accessed real feedback (for many, it was the first time)
CTO mandated roadmap alignment with top DevX themes
19 platform teams committed roadmap actions based on their feedback
BU leaders incorporated data into annual goal-setting
PMs built OKRs grounded in evidence
Several teams secured additional budget by leveraging data from DevX dashboards to build a business case
Increased requests for deepdive research
For the Firm - Culture & Strategy
Eliminated the need for 28+ separate team-run surveys, preventing bias and survey fatigue
Engineering leaders began seeking deeper research partnerships
Enabled “We heard you” communications as part of a closed-loop strategy
Improved experience for engineers (eg. CI/CD, onboarding, entitlements, documentation, plus many more)
Supported rollout of GSat experience measurement framework
Increased research visibility at senior levels
Legacy & Reflection
This work didn’t just produce dashboards. It changed how the organisation understood and acted on the voice of its engineers.
Leaders became accountable to the feedback they could now see directly. Roadmaps shifted from assumption-based to evidence-informed.
And research earned a seat at the strategic table by enabling, not controlling, access to insight.
