GMV Masks True Platform Economics

Diving deeper into

Jemma White, COO of Prolific, on why humans ensure AI safety

Interview
gross revenue or GMV includes everything that’s paid out to contractors and participants
Analyzed 6 sources

Headline revenue in human data often overstates the actual economics, because a large share of each dollar passes straight through to the people doing the work. In practice, many vendors count participant pay, contractor pay, and delivery labor inside topline figures, even though the durable business is the smaller take rate left after payouts. That is why net revenue and gross margin say more about quality, pricing power, and staying power than GMV does.

  • Prolific runs more like a marketplace than a labor heavy BPO. Researchers set eligibility, sample size, and pay, participants get roughly 70% of study budgets, and Prolific keeps the remainder as platform revenue. That structure makes the gap between GMV and net revenue especially important when comparing it with peers.
  • The contrast is clear in adjacent players. Handshake AI is estimated to send 60% to 70% of spend to contractors, leaving 25% to 40% gross margins, while Invisible charges roughly $30 to $45 per hour and pays raters $15 to $20, showing how much of topline can be labor pass through rather than software economics.
  • Scale sits between software and outsourcing. It bundles tooling with a large contractor base and has reported 50%+ gross margin, but its model still includes recurring human labor costs on every task. That makes topline growth less informative than how much automation and pricing discipline lift retained revenue over time.

The market is moving toward a cleaner split between labor volume and true platform value. As buyers become more sophisticated, companies with transparent pricing, repeatable self serve workflows, and higher retained margins should separate from vendors whose growth is mostly a bigger payroll flow disguised as revenue.