Industries / Clinical trials
Clinical trials · sponsors, CROs and site networks

Recruitment that learns from every enrollment.

The biggest leak in trial recruitment isn't ad spend, it's the form fill. A patient submits interest, then waits in a callback queue, and most never answer the phone. We replace that handoff with instant qualification and feed every enrollment back to your ad platforms, on a HIPAA-compliant system.

HIPAA-compliantMetaGoogle AdsProgrammaticSite networksHashed signals only
Study recruitment · last 30 daysIllustrative example
Interested
1,240
Prescreen passed
520
Site visit
168
Randomized
41
FrogFish Signal now
ACTIONSite 014 has 22 prescreen-qualified patients waiting 3+ days for a call.
→ Patients re-routed to Site 009 · reminder texts sent
The problem

Most recruitment dies between the form and the phone call.

Platforms optimize to form fills. Sites call back hours or days later. By then, the patient has moved on.

01

The callback gap.

Interest peaks the moment a patient clicks. Every hour before the first contact costs enrollments.

02

Platforms can't see enrollment.

Screen fails and randomizations never reach the ad platforms, so they keep finding more people who won't qualify.

03

PHI makes feedback hard.

Most teams avoid sending any signal back at all, so the algorithms never learn.

Offline informs online

The loop, for study teams.

What happens after the lead, in calls, consults and contracts, goes back into how we bid, target and create. Every turn makes the next enrolled patient cheaper.

Offline data we pull

Prescreen completions, eligibility, site visits and randomization status, passed as anonymized, hashed signals. No PHI leaves the secure boundary.

What changes online

Meta, Google and programmatic stop optimizing to form fills and start optimizing to qualified patients. Lower cost per enrolled patient, higher screen-to-randomize rates.

What your team gets

Instant prescreening when intent is highest, qualified patients routed to the right site immediately, and alerts when a site's queue backs up.

Systems we connect

Prescreeners and site CRMs, Meta, Google Ads, programmatic, with HIPAA-compliant data handling end to end.

We optimize toCost per randomized patient
How we operate

Three connected loops. One flywheel.

Measurement feeds targeting. Targeting feeds qualification. Qualification feeds measurement. Each loop strengthens the next.

01 · Measurement

Lure™

Our analytics layer captures enrollment data and passes it to ad platforms as compliant, anonymized signal.

02 · Targeting

Smarter signals

Platforms learn which leads enrolled and optimize toward qualified patients.

03 · Qualification

Strike™

Patients prescreen in real time; contact rates jump 50–70% versus form-fill funnels.

04 · Routing

Site matching

Qualified patients go straight to the site with capacity.

05 · Reporting

Study dashboards

Cost per randomized patient by site, indication and channel.

The system, running

Sites get patients, not form fills.

The system watches every study's funnel daily and flags the bottleneck, whether that's a site queue, a channel or a screener question, with the fix ready to approve.

  • Site backlog and no-contact alerts
  • Screener drop-off flagged by question
  • Weekly cost per randomized patient by channel
# study-recruitmentIllustrative example
FrogFish Signal 9:05 AM
BACKLOGSite 014: 22 qualified patients uncontacted for 3+ days.
→ Re-routed to Site 009 · reminders sent
FrogFish Signal 1:30 PM
SCREENERQuestion 6 (medication history) drives 38% of prescreen drop-off.
→ Plain-language rewrite ready for sponsor review
FrogFish Signal Mon 8:00 AM
WEEKLY41 randomized this month. Meta lookalikes from enrolled patients beat interest targeting by 2x.
What this means for you

Built for teams that need to enroll fast.

Single-indication studies or portfolios across therapeutic areas. Pressure-tested in oncology, neurology, rare disease and broader categories.

Sponsors

Faster enrollment, lower cost per patient

Cost per patient drops as the platforms learn what qualified looks like.

CROs

Margin-friendly recruitment

Media, screening and analytics under one accountable partner.

Sites

Pre-qualified patient flow

Sites receive patients ready to screen, not random form-fillers.

Proof

Recruitment run by people who've done it at Alphabet scale.

Founder Alex Fleshner spent 13 years at Google, including leading the Google Ads lead-gen vertical, Head of Media Strategy at Verily and AI Transformation Lead at Google Cloud AI. Senior people run every account.

50–70%Higher patient contact rates with real-time prescreening versus traditional form-fill funnels
0 PHILeaves the secure boundary. Only anonymized, hashed signals go to ad platforms
13 yrsAt Google, including Head of Media Strategy at Verily, Alphabet's health company
FAQ

Questions study teams ask us.

Is your patient recruitment HIPAA-compliant?

Yes. Enrollment signals are anonymized and hashed before they reach any ad platform, and no protected health information leaves the secure boundary.

How do you reduce cost per randomized patient?

We feed prescreen, site visit and randomization outcomes back to the ad platforms so they optimize toward patients who qualify, and we prescreen in real time so qualified patients don't drop out while waiting for a callback.

Do you work with CROs as well as sponsors?

Yes. We work with sponsors, CROs and site networks, and can bundle media, screening and analytics under one partner.

Which therapeutic areas have you recruited for?

The system has been used in oncology, neurology, rare disease and broader therapeutic categories.

Not another agency

Have an upcoming study?

Book a 30-minute call. We'll review your recruitment funnel and indication, and tell you whether our system fits.

Book a recruitment review →
Product screens show illustrative examples, not client data. Results shown are from FrogFish client engagements.
Get a Free Audit →
Services Industries AI Agents Paid Media AI Search About Contact Book a call →