Participant selection controls who gets into your study: the system (Automated) or you (Hand picking). You'll find it on the Recruit tab under Settings, as Participant selection.
How automated works
The system invites top-scoring applicants automatically, either straight into your task or into available session slots. Scoring comes entirely from your multiple-choice screener options marked approve or reject. Two things follow from that:
Open-text screener answers carry no weight. The system can't score them, so an open-text-only screener gives automated mode nothing to work with.
Screeners rank, they don't block. Anyone can apply; lower-fit applicants just score lower. Make your must-have criteria explicit multiple-choice questions with reject options.
See how screeners work before choosing automated.
When to choose automated
Speed and low effort. It suits early discovery calls, large-scale unmoderated studies and surveys, and any study where your screeners fully define a good participant and the demographic mix doesn't matter.
Your remaining job is to review submissions within 24 hours for unmoderated studies, or simply turn up to the sessions for moderated ones.
When to choose hand picking
Control. You review each applicant's screener answers and invite the people you want. Choose it when the mix matters (quotas across groups), when fit is nuanced, or when you want to read open-text answers before inviting.
It costs more of your time, but two features take most of the work out of it:
Quota targets on the Recruit tab let you define audience segments with custom quotas.
Segments let you filter and bulk-invite applicants once the study is live.
A simple split? Consider two automated studies
If your groups differ on a single criterion, such as gender or one screener answer, you can run two automated studies, one per group. For anything more nuanced, use hand picking with quota targets or segments.