A marketplace has two products in one interface. Sellers need to list quickly, buyers need to trust fast, and the same page often serves both badly.
Testers can be assigned each role separately so you see where each side stalls.
What you get: measured facts separated from human judgement.
Start an auditWhat we check
Supply-side onboarding
How long it takes to publish a first listing and how many fields feel unjustified.
Search and filtering
Whether a buyer with a specific need can reach a relevant result in a few actions.
Trust between strangers
Ratings, verification, dispute policy and payment safety, judged as a first-time user sees them.
Empty categories
Liquidity gaps that make the platform look abandoned, and how the interface handles them.
First transaction
Testers complete or attempt a transaction and report every point of hesitation.
Typical findings
- Listing form asking for details a seller does not have yet.
- Filters that return zero results with no fallback suggestion.
- No visible dispute or refund policy before payment.
- Buyer and seller messaging hidden behind an account wall.
Questions
Can testers play both roles?
Yes, create two tasks with different briefs and compare the reports.
Do you check the payment provider?
We check the experience, not the provider's internals.
How is severity decided?
By impact on completing the main action, with evidence attached to each finding.
See also
A roast is an automated audit plus reports from real testers, with evidence attached.