Microsoft · State of California · Research operations

Building the machinery that makes research scale

Research only scales on the infrastructure most people never see — panels, recruiting, consent, incentives. Building it is as much a program-management job as a research one, and I've built it from the ground up in two very different organizations.

Role
Research & program leadership
Focus
Research operations · Participant panels · PM
Tooling
Two custom in-house tools · UserTesting · Qualtrics · Ethnio
2
organizations, built from zero
8,000
participants · California panel
2
custom in-house tools
50%
faster recruiting · contact to accept
The problem

Insight doesn't scale. Infrastructure does.

A single study is a craft project. A research practice — one a whole organization can rely on, again and again, without reinventing recruiting every time — is an operations problem. Someone has to build the panels, the screeners, the consent and incentive systems, the tooling and the vendor relationships that sit beneath the work. Do it well and it disappears into the background; skip it and every team quietly pays the tax, one slow, expensive study at a time.

The work

Two organizations, two operations built from zero

The same discipline, in two very different settings — one enterprise, one public-sector — each starting from nothing.

Microsoft

An in-house panel, on tools I built

Rather than pay to re-recruit the same kinds of people study after study, I built two custom in-house tools — one for recruiting, one for managing the panel — behind an in-house research panel drawn from a mix of sources. It gave product teams a reusable pipeline instead of starting cold every time, and cut recruiting time from first contact to accept by 50%.

State of California · ODI

A state research practice from scratch

At the Office of Data & Innovation there was no research-operations function at all. I built one: an 8,000-participant panel with screeners, consent, and incentive systems across UserTesting, Qualtrics, and Ethnio — designed to reach residents a public service can't afford to miss, and to feed the state's AI evaluation program at scale.

The point

Researcher and operator, in one seat

This is the part of the job that rarely makes the highlight reel: the tooling, contracts, and systems beneath the insight. It's also where a researcher who can run a program earns their keep — turning research from a string of one-offs into a capability an organization can count on. Having built it twice, in an enterprise and a government, is the clearest evidence I can offer that I operate as much as I research.

Insight is the visible half of research. The other half is operations — and it's the half that decides whether the insight ever arrives. — how I think about research ops

Happy to walk through how I built these.