Get startedFeFe Assurance
When you commission human data, you deserve human data. FeFe Assurance is our five layer system for keeping every response on the platform genuine, considered and worth building decisions on.
Every participant completes a full health screening, an attention check, a written response that shows genuine effort, and an authenticity pledge covering AI use and duplicate accounts. Fail the check twice and the account can never take a study.
Verification does not stop at sign up. Participants re-confirm their profile every six months, and accounts that stop meeting our standards are suspended from taking studies.
Completion codes are matched server side, so mismatches are flagged for you. Time on task is measured against your own estimate, and exceptionally fast submissions, under 40% of your stated length, are flagged and can be auto-rejected. Every tool is included free.
Each participant carries a live quality record of approvals and rejections. You can screen for a minimum approval rate and a minimum number of approved studies, and participants with too many rejections are removed from the pool.
Good participants are protected as firmly as poor ones are removed. FeFe enforces a fair pay floor of £6 per hour, pays 100% of the reward to the participant, and auto-approves unreviewed submissions after 21 days so nobody waits on their money.
These are three different problems and each needs a different defence. Here is how FeFe handles all three, at sign up, in the pool and inside your study.
Scripted accounts give themselves away with inhuman speed and repetitive patterns. FeFe blocks them with a required email confirmation, an attention check, a written response that has to show real effort, and a rate limit that stops any account grabbing more than three studies in a minute.
Agents that reason and adapt are harder to spot, so we do not rely on one test. Access depends on a completed human health screening, a single account per person, a six monthly profile re-confirmation, and a live approval record that has to be earned study by study.
The most common threat today is a real person pasting your open questions into a chatbot. Every participant signs an authenticity pledge, every study states its AI policy up front, and you can report a submission as AI written in one click. Two confirmed reports remove that person from the pool.
Set your AI policy when you create a study: no AI assistance, AI allowed if the participant discloses it, or AI allowed. Participants see it before they take a place.
Every study on FeFe runs under all five layers. There are no quality add-ons and no premium tiers for basic integrity. If a submission is rejected, the place is released straight back to the pool so your sample stays full.
Run a study on FeFe