Conversion tracking for ClinicalTrials.gov referrals
Your coordinator's phone rings, the caller "found the study online," and nobody can say whether that was your Google ad or the ClinicalTrials.gov record. The registry is a federal listing, not a lead platform: a patient searches by condition and location, opens your record, and calls the number under Contacts and Locations. Nothing reports that call to the ad accounts you pay, and nothing flags the patient who clicked your ad, checked the NCT number, then called from the registry. We split registry calls from ad calls at intake and send only real consents back to Google and Meta.
Free30 minutesNo obligation
ClinicalTrials.gov · what breaks
What goes wrong with ClinicalTrials.gov referrals
- 01
Every caller lands on the same coordinator line
The record shows one phone number and one email per site. If the same ones sit in your Google and Meta ads, registry and ad patients arrive mixed: you pay the platforms for calls the registry sent free and lose credit for the ad calls that were real.
- 02
The ad gets no credit when the patient checks the registry first
A cancer center advertises a study; the patient sees the ad, looks up the NCT number to confirm it is real, and calls from the record instead. The click shows no conversion, and Smart Bidding, which learns only from the conversions you report, moves spend toward keywords that enroll fewer people.
- 03
Pre-screen calls get counted as wins
Registry callers who fail eligibility still count as conversions if a phone call is the conversion. The platform then optimizes for callers, not consents, and your cost per randomized participant climbs while the ad reports look fine.
- 04
A stale site status pulls in calls for a closed cohort
Responsible Parties have 30 days to update a site's status in PRS. A site still marked Recruiting after enrollment closes keeps generating calls that look like leads and burn coordinator hours on a cohort that no longer exists.
ClinicalTrials.gov · the fix
What changes — ClinicalTrials.gov
Registry referrals become their own channel
A dedicated line and inbox on each study record, so a registry call is logged as one before anyone picks up.
Enrollments go back to the ads, not phone rings
Consents are matched to the original click and uploaded from your enrollment records, with pre-screen failures left out.
One report per study, per source
Cost per pre-screen, screening visit and consented participant, split by ClinicalTrials.gov, Google Ads and Meta.
ClinicalTrials.gov · architecture
What ClinicalTrials.gov actually lets you do
An NIH registry with a read-only public API. No pixel, no tag container, no webhook, and only the Responsible Party can edit the record, so every hook below lives on your side of the phone call.
- 1
Contacts and Locations on the study record
A Study Contact plus a per-location Contact with name, phone and email, shown only while the study is Recruiting or Not yet recruiting. This is where a tracking number and dedicated inbox go, within your sponsor's and IRB's rules.
- 2
Labeled links under More Information
Records can carry labeled URLs; we have seen a sponsor point one at its own site-finder form. A link to your domain with a registry source tag is the only place a web visit from the registry can be measured.
- 3
Read-only public API
Version 2 is GET-only with no key and returns records, per-location status and listed contacts as JSON or CSV. Good for auditing what the public sees about your sites, useless for sending data anywhere.
- 4
Deep links to the record
Study pages sit at a stable address keyed on the NCT number, with a #contacts-and-locations anchor. If an ad must send patients to the record itself, the click is the last thing you can measure until the phone rings.
- 5
No analytics, no conversion tools
Nothing in the registry's documentation lets a site measure views, calls or clicks on its record. Counting happens on the phone line, the inbox and your own study page.
ClinicalTrials.gov · diagnose
Where ClinicalTrials.gov referral tracking breaks, in diagnostic order
We start with the phone number, because on registry traffic the usual fault is two sources sharing one line, not a missing tag.
- 01
Compare the number on the record with the number on your ads
If they match, registry and ad calls are already indistinguishable in every report you have.
- 02
Check whether the pre-screen script asks how the patient found the study
And whether the answer is stored as a field, not a note.
- 03
Check each location's recruitment status
Against what the site is really doing. The record only changes when the Responsible Party updates PRS, and stale statuses create calls no campaign should be credited for.
- 04
Trace one consented participant backward
CTMS entry to pre-screen to first contact. If the chain breaks at the call, consents cannot reach the ad platforms.
- 05
Look at what the platforms call a conversion today
If it is a call or a form, pre-screen failures are inflating the count.
ClinicalTrials.gov · prove
The QA evidence — ClinicalTrials.gov
Run on live records and real coordinator calls, study by study, not on a test form.
Source of truth
Your CTMS or enrollment log, by study and site: pre-screens with outcome, screening visits and consents by date
- One test call to each active record's number, logged under the registry source with the right NCT number.
- One visit through the record's link to your study page, confirmed as a registry-tagged session.
- A closed month of consents matched one by one to ad clicks, registry calls or neither.
- Pre-screen failures confirmed absent from every uploaded conversion.
- A withdrawal reversed where the platform supports it.
- Repeated 30 days later and after any site status change on the record.
ClinicalTrials.gov · free audit
What we check in the free audit
- Compare the number on the record with the number on your ads
- Check whether the pre-screen script asks how the patient found the study
- Check each location's recruitment status
- Trace one consented participant backward
- Look at what the platforms call a conversion today
Checked against
Your CTMS or enrollment log, by study and site: pre-screens with outcome, screening visits and consents by date
Free30 minutesNo obligation
ClinicalTrials.gov · when not to buy
When you don't need this for ClinicalTrials.gov
Some research sites get everything they need from a dedicated extension and a spreadsheet.
- If you run no paid recruitment ads, a separate extension on the record and a column in your enrollment log counts registry referrals fine.
- If your sponsor or CRO runs central recruitment and you only field hand-offs, the measurement belongs with them.
- If one coordinator runs one study and already logs the source of every call, we would only be formalizing what works.
- If a site status is simply out of date, a PRS update by whoever holds the record is a far smaller job than any tracking build, and we will tell you to do that first.
ClinicalTrials.gov · what we deliver
The event map for ClinicalTrials.gov referrals
Registry contact received
A call or email to the contact on your study record, tagged with the NCT number and site.
Ad-assisted registry contact
A registry caller who clicked an ad first, found by the intake question or by matching their number to a tagged study-page visit.
Pre-screen completed
The eligibility call, with outcome recorded. Ineligible callers are logged, never sent to an ad platform.
Screening visit scheduled
The first qualified step, sent as a secondary conversion so bidding gets an early signal.
Consented or randomized
The primary conversion, uploaded from your CTMS or enrollment log, valued per study if you set a figure.
Screen failure or withdrawal
Retracted in Google Ads through a conversion adjustment, left out of the Meta upload, excluded from cost per enrollment.
FAQ
ClinicalTrials.gov questions, answered
No. The registry publishes your record on its own domain and documents no tags or analytics for sponsors. The measurable points are the phone number and email on the record and any link to a page you control.
Give the record its own number and inbox, and add a source question to the pre-screen script. Matching the caller's number against tagged visits to your study page then catches people who clicked an ad and called from the registry.
Only when the patient started from an ad. A pure registry referral has no click to attach to, so it stays in your own reporting as a registry enrollment. Ad-assisted consents go back to the platform from your enrollment log.
Google is counting calls and your budget counts consents. Most callers fail pre-screening. Moving the primary conversion to consent, uploaded from your records, lines the two numbers up.
Conversions go out as a click ID or a SHA-256 hashed email or phone with a study-level label, never a condition name or anything from the patient record. We confirm the exact fields with your IRB or compliance lead before anything leaves your system.
More on ClinicalTrials.gov
- Enhanced ConversionsEnhanced Conversions enrich the conversions that already fire. First we make sur…
- Meta CAPIA one-click CAPI integration sends thin events. That's why your Event Match Qual…
- Cross-Domain TrackingYour checkout provider is the best performing traffic source in your GA4 account…
RELATED GUIDES
Booking tools beside ClinicalTrials.gov, and what their bookings feed
- ElderLawAnswersYour ElderLawAnswers leads are being billed to your ad campaigns
- LendingTreeYour ads are getting credit for loans LendingTree matched to you. We separate the two.
- ThomasnetIs it Thomasnet or Google Ads that's winning your RFQs?
- Alarm.comAd reports look great but installs don't match? Check your Alarm.com referrals.
- Best LawyersBest Lawyers conversion tracking for firms that can't tell what their recognition actually signs
- Dude Ranchers' AssociationDude Ranchers' Association guests your reports hand to Google Ads
- LegalZoomYour LegalZoom consults turn into clients, and Google Ads takes the credit
- Apartments.comApartments.com leads are padding your Google Ads results
Free · 30 min
Get ClinicalTrials.gov referral tracking right
Free30 minutesNo obligation
