Case
studies
Six featured deep-dives across LinkedIn, Eventbrite, Observable and Airbnb, plus a research archive of two further studies.
Four deep dives.
Passcode required · contact for access · all outcome figures gated
Premium Duo, Phase 1
De-risking a paired-subscription launch with targeted qualitative research running in parallel with a 15-week live experiment.
- Role
- Lead UXR · 1 of 2 researchers on Premium acquisition
- Method
- 8 generative interviews, parallel to a 15-week test
- Timeline
- Dec 2024 → Mar 2025 · 15 weeks
How qualitative signal de-risked a multi-million-dollar launch decision when the experiment alone couldn't say why members were buying.
Premium App Explorations: Five Principles to One Feature
A six-round programme that built a framework for what makes a subscription feel premium, tested it round by round against a design exploration, then narrowed it to one feature.
- Role
- Lead UXR · Premium App Explore
- Method
- Unmoderated study (n=25) + co-design workshop (n=6) + 3 concept tests + behavioral simulation (13 sessions)
- Timeline
- Apr → Dec 2025 · 6 rounds + follow-on
How six rounds of member research suggested that design alone doesn't make Premium feel premium, and ended up cited by name in Premium's FY26 OKRs and H2 plan.
Premium Acquisition Survey
A two-population read on the 50% bounce problem, pairing a two-million-member behavioral analysis with depth interviews recruited from the same funnel.
- Role
- Lead UXR · UAM Acquisition
- Method
- Behavioral analysis (2M+ members) + 9 depth interviews
- Timeline
- Jan → Apr 2026
How completers and skippers turned out to be two populations, and what a dual-path intent survey would mean for the funnel.
Pricing & Packaging, Eventbrite
An eight-month strategic research program that informed Eventbrite's shift to a tiered subscription model for creators, from packaging-strategy development through 2.1 and 2.2 rollout.
- Role
- Principal UXR · Pricing & Packaging program lead
- Method
- Mixed-methods: qual + behavioral analytics + moderated and unmoderated user testing
- Timeline
- Nov 2022 → Nov 2023 · 12 months
How a recognition-vs-recall cognitive-load framing reshaped where pricing showed up in a new creator's first session.
Observable, Audience & Jobs-to-be-Done
Testing whether a developer platform had really inherited its founder's community, then building and validating a jobs-to-be-done model against the answer.
- Role
- Founding and only researcher
- Method
- Community survey → cohort analysis → 24 interviews → validation survey
- Timeline
- Oct 2020 → Dec 2021
How an assumption about who used the platform survived until someone tested it, and what the correction revealed about who was really building.
Continuous Delivery, Post-Launch
A post-launch study of Airbnb's continuous delivery platform, pairing a population survey with moderated testing to explain a gap neither method could explain alone.
- Role
- Experience Researcher · Developer Platform · survey co-run with the platform team
- Method
- Qualitative survey + 6-participant moderated usability testing
- Timeline
- Q1 2020
How 91% of engineers understood a safety feature, 81% never switched it on, and watching them work explained why.
From the archive.
Contact for access to full source material
Long-term Daily Active Use (DAU)
Mixed-methods study (RPIA + qualitative) that reframed how Premium thought about daily-active behavior, informing Build 2.0 feed personalization and email notification strategy.
Testability: The Human Element
Qualitative IDI study across Airbnb's four primary code repositories surfacing how developer testing culture, build-time friction, and tooling gaps shape what gets tested and what doesn't.