How we test and review survey sites

We open a standard member account on each panel we write about — no press access, no seeded profile. We then run at least ten survey sessions across a minimum of 30 days, because new-account invitation quality is not representative.

Sessions are run on ordinary residential connections without VPNs, on both a laptop and a phone, matching how members actually participate.

Per session: invitations received, surveys started, completes, screen-outs, minutes spent, rewards credited, and the gap between stated and actual length of interview.

Per panel: cash-out threshold, time from request to receipt, whether screen-outs are compensated, and the response time and usefulness of one support ticket we always file.

We complete at least one real cash-out before publishing timing claims. A panel we could not withdraw from is reported as such.

Some links on this site earn us a commission. Commission never determines inclusion, ordering or wording, and we publish criticism of panels we are paid to refer. Where a panel failed our checks, we say so and do not link it.

Panels change terms frequently. Every guide carries an updated date, and we revisit payout figures at least twice a year. If you can show a figure is wrong, send the evidence through our contact page and we will correct it and note the change.