Building the GCP Marketplace roadmap from zero, one partner interview at a time

Overview

GCP Marketplace's transaction platform MVP was effectively a "what should exist, and in what order" problem with no existing user research foundation to build on — happening as GCP's UXR org scaled from roughly 40 to 400+ researchers across 15+ products.

Rather than validate a set of assumptions, the goal was to build the roadmap's foundation from the ground up: understand partner needs first, then let prioritization follow from evidence.

Approach

Understood the space first: reviewed PRDs, wrote user scenarios to illustrate concepts, then worked with leads to split research needs into P0 (inform the North Star) and P1 (validate MVP scope and requirements).

Built an 8-page research protocol spanning multiple altitudes, from high-level needs down to implementation detail, then worked with business development and field teams to identify top partners and consolidate what was already known before collecting anything new.

Ran 3 interviews per partner organization — about 30 hours of interviews total — summarizing after each one rather than batching analysis to the end.

Analyzed with two coding approaches matched to question type: systematic coding for semi-structured sections, grounded coding for open-ended ones. Produced an 18-page report that validated parts of the MVP, recommended changes to 2 launch features, and surfaced 42 missing post-MVP features (all later included in the release), backed by dozens of grounding quotes.

Prioritized the 42 missing features down to the top 25 with a MaxDiff survey, working with business development to reach 36 of 68 targeted partners — then ran usability studies on UX design mocks for the top 5 features to shape exactly how they should be delivered.

Closed the loop by establishing user-driven decision-making as an ongoing cadence rather than a one-off study: a recurring feature-prioritization exercise with customers and partners, plus standing feedback channels (in-product surveys, usage metrics).

Impact

First launch represented roughly 25% of 2020's total Marketplace revenue within 6 months, with only a handful of partners onboarded.

60% reduction in reported customer issues.

Grounded the roadmap for a team of 25 engineers and 3 PMs directly in user needs; research is leveraged in every subsequent PRD for the team.

Normalized gathering feedback early in product development going forward, rather than validating decisions after the fact.

"Yann, thank you for navigating a complex research effort... these efforts will set the foundation for the next generation of pricing design on the marketplace." — PM Lead, Marketplace transaction platform.