Reading Your Recruitment Funnel: Where Are You Actually Losing People?

"We need more leads" is the most common self-diagnosis at under-enrolling sites — and often wrong. The funnel usually knows better.

Map the stages

A minimal recruitment funnel: lead → contacted → pre-screen passed → screening scheduled → screened → enrolled. Count people at each stage for a defined cohort (say, all leads from March) and compute stage-to-stage conversion. Cohorts matter: mixing last week's uncontacted leads with January's closed ones muddies every rate.

Diagnose by drop-off pattern

Lead → contacted below ~60%: an operations problem. Speed-to-contact, attempt counts, and calling times are the levers — not ad spend.

Contacted → pre-screen pass unusually low: a targeting problem. The channel or the ad promise is attracting the wrong population. Fix the message before scaling the budget.

Pass → scheduled weak: a friction or motivation problem. Scheduling requires too many steps, slots don't fit working people, or the study's burden is landing badly in the conversation.

Scheduled → screened low: a no-show problem with a known playbook (reminders, reschedule links, expectation-setting).

Screened → enrolled low: protocol reality or consent-stage surprises. Feed screen-fail reasons back into pre-screening; feed consent withdrawals back into recruitment messaging.

Benchmarks are local

Published conversion benchmarks vary wildly by therapeutic area and channel mix; your own trend line is the useful comparator. The question isn't "is 12% lead-to-enrollment good?" but "why did it fall from 15% last quarter?" — and the stage-level view answers it.

The discipline

One funnel per study, refreshed weekly, reviewed at the same standing meeting as your other metrics. Sites that do this stop having arguments about whose fault enrollment is. The funnel points; the team fixes the stage it points at. More leads is the last resort — and, pleasingly often, unnecessary.