Research Operations

The team did not need more research craft. It needed a system that could move customer evidence at product speed.

2× research panelMonths → <2 weeksPower BISelf-service research

The constraint

Research demand was growing faster than the team could deliver it.

Recruitment could take months. Study setup was repeated by hand. Insights were difficult to reuse. Even strong research arrived too late for some product decisions.

The bottleneck was not research quality. It was throughput.

What I built

I treated research like an operating system, not a queue of projects.

Standing participant panel

Expanded the customer panel and made recruitment a reusable capability instead of a new project every time.

Standard workflows

Created repeatable processes for screening, scheduling, consent, incentives, and study setup.

Shared evidence

Used a research repository and Power BI to make customer evidence easier to find and combine with product data.

Teach-to-Fish

Built training, templates, and guidance so product teams could run appropriate research without waiting for a specialist.

Scale

The system changed which research was possible.

customer research panel
<2 weeksrecruitment lead time
Cross-teamself-service enablement

Faster recruitment did more than save time. It allowed research to fit inside product decision windows instead of arriving after the decision.

What I learned

Scaling judgment means designing the system around it.

The same pattern appears in my newer AI evaluation work: high-volume inputs, a triage layer, human judgment where it matters, and tooling that makes the judgment sustainable.

What this demonstrates

Program management, operating-model design, cross-functional enablement, analytics, process automation, and the ability to scale a specialist capability across an organization.