Skip to main content
← Back to selected work

02 / Marketplace

A sourcing marketplace that explains itself

Mixed-method research into a rebuilt onboarding and pricing model.

Case study cover

Context

B2B sourcing / marketplace

Series B marketplace startup, ~60 employees

Company name withheld under NDA.

Role

Me — Mixed-method research, onboarding redesign, pricing model UX

Team — 1 PM, 1 growth engineer, 2 product engineers

The constraint

First-time buyers were dropping off mid-onboarding at a rate the team couldn't explain from analytics alone — the funnel data showed where people left, not why. Pricing itself was also a black box: quotes depended on supplier-side variables buyers never saw.

What we tried

01

Tried — Redesign the onboarding flow first, based on the funnel drop-off data, before doing new research.

Learned — Early prototype tests against the funnel-informed redesign still failed for the same reasons as the original — the funnel told us where people left, not why, and we'd just designed a different version of the same problem.

Did — Paused the redesign and ran 14 moderated sessions with recent drop-offs specifically, which surfaced that buyers didn't trust a quote they couldn't explain to their own finance team.

02

Tried — Add a tooltip-based FAQ layer explaining pricing terms, to address the trust gap cheaply.

Learned — Buyers weren't confused about terminology — they needed to justify the number internally, and a tooltip doesn't produce something you can forward to a manager.

Did — Designed an itemized quote breakdown buyers could export, showing exactly which supplier-side factors drove the price.

03

Tried — Launch the itemized breakdown to all buyers at once.

Learned — Testing against the full buyer base risked masking a bad interaction for a specific segment inside an averaged result.

Did — Rolled out to first-time buyers only first — the segment the research targeted — before expanding once the metric held up.

Outcome

+31%

First-time buyers completing onboarding rose 31%, and the itemized quote breakdown became the most-exported artifact in the product within two months of launch.

Arun Jolly Joseph