TL;DR
- FTA Global’s life sciences research team tested 420 buyer questions across ChatGPT, Gemini and Google AI Mode to understand CRO discovery in the US market.
- Buyer role, clinical requirements and buying stage influenced which CROs appeared in AI answers.
- Eight of the top ten CROs at the long list stage remained in the top ten at the final RFP stage.
- CRO and vendor websites accounted for 59% of cited links, making service pages a priority for content improvement.
- CRO marketing teams should assess visibility across relevant buying scenarios and review the sources shaping how AI presents their firm.
What did FTA’s US CRO study measure?
We developed 420 questions in US English across seven buyer roles and five buying stages. Roles covered founders, chief medical officers, clinical operations, regulatory affairs, clinical pharmacology, biometrics and safety, and procurement.
Questions followed the journey from learning to building a long list, shortlisting, checking risk and preparing an RFP. Examples covered oncology partners for smaller biotechs, Phase 1 capabilities and concerns before signing with a named CRO.
Collected on 15 October 2026, the research produced 2,953 usable answers and 17,278 cited links. Company naming questions were repeated three times per platform.
The questions modelled buyer situations. The study examined generated answers, rather than observing individual pharma or biotech sponsors making purchases.
ChatGPT and Gemini were accessed through APIs with search enabled. Google AI Mode responses were collected through DataForSEO. Results represent one day of testing and may differ from personalised app experiences. US CRO Study(1).pdf
Which CROs appeared most often in AI answers?
Medpace led with a 48.6% mention rate, followed by IQVIA and Worldwide Clinical Trials.
Mention rate measures the proportion of eligible answers naming a provider. The denominator was 1,387 company-naming answers containing at least one provider, rather than every answer collected.
Separately, the five most mentioned providers accounted for 32% of all provider mentions within those answers. Of 600 providers named in that subset, 459 appeared fewer than five times.
For marketing teams, occasional inclusion can conceal weak coverage. Track whether your CRO appears consistently in the services and buyer situations that matter commercially.
How do buyer role and company size change the shortlist?
When questions identified the buyer as a small biotech, midsize CROs received 43% of classified provider mentions, compared with 27% for other buyers.
Large CROs still appeared first 40% of the time in small biotech scenarios. Positioning around smaller pharma and biotech sponsors therefore needs evidence of fit, including relevant delivery experience and how accounts are supported.
Buyer role produced another clear difference. Specialists received 60% of mentions in questions from heads of clinical pharmacology, compared with 11% for procurement.
Certara, Metrum Research Group and Allucent led the clinical pharmacology buyer list. Procurement’s leading names were IQVIA, ICON and Worldwide Clinical Trials.
A single message about comprehensive capabilities will leave important questions unanswered. Clinical leaders need specific technical evidence, while procurement content should address vendor evaluation and commercial concerns.
Why does visibility before the RFP matter?
Eight of the ten leading CROs at the long list stage remained in the top ten at the final RFP stage.
The pattern suggests that early discovery deserves sustained attention. Content explaining suitable partners, service scope and therapeutic experience can support the first company naming questions.
Risk checks also changed visibility. Medpace and Worldwide Clinical Trials appeared more frequently at that stage, while Veristat’s mention rate fell substantially.
Being named during a risk discussion does not establish a positive recommendation. Teams should review what the answer says alongside how often their brand appears.
The RFP comparison contained only 64 company naming answers. It indicates a pattern in the study, but it doesn't show how actual contracts were awarded.
Can one AI platform represent your CRO’s visibility?
ChatGPT and Gemini shared an average of only two names among their top five for the same question. ChatGPT and Google AI Mode shared 1.8, while Gemini and Google AI Mode shared 1.6.
Their aggregate leaders also differed. Worldwide Clinical Trials led ChatGPT, IQVIA led Gemini, and Parexel led Google AI Mode.
Repeated testing revealed further variation. Across three runs, the same provider remained first only 39% of the time. The same five names appeared, in any order, just 3% of the time.
Location offered less differentiation in this study. For ChatGPT, changing the city changed 2.1 of the top five names on average, compared with 2.2 when repeating the question from the same place.
The five location tests covered US biotech hubs on ChatGPT and Google AI Mode. Results did not establish a city effect beyond ordinary answer variation.
Where can specialist CROs build stronger Context Authority?
In this study, Context Authority is a score from 0 to 100 that measures how often a company appears within a service or disease area, with equal weighting across buyer roles and stages.
Cytel led biostatistics and programming with 80. Certara led clinical pharmacology and modelling with 78. Worldwide Clinical Trials led neuroscience with 78, while Medpace led five of seven disease areas.
The report marked comparisons with fewer than 60 answers as small samples and warrants cautious interpretation.
The opportunity is specific. Build evidence around the clinical work your organisation performs particularly well, then assess whether AI associates your firm with that work.
FTA’s Context Authority explores why buyer context belongs in visibility measurement. Here, the service and disease comparisons show how an overall score can hide areas of strength.
Which sources should CRO marketing teams prioritise?
CRO and vendor websites accounted for 59% of all cited links. Trade and business news represented 4%.
Among ChatGPT and Google AI Mode links to CRO and vendor websites, 49% pointed to service pages. We excluded Gemini from that page-type analysis because its results provided website names rather than comparable page details.
Review service pages for clear scope, therapeutic experience, delivery capabilities and substantiated examples. Give buyers useful answers that can also stand on their own when cited.
For example, an oncology service page should explain the trial phases supported, relevant study experience and how the team approaches delivery.
A clinical pharmacology page needs a different level of detail about modelling capabilities and specialist expertise. Let the buyer’s actual requirement shape the content.
Reputation checks need separate attention. When asked about 40 CROs by name, AI raised ownership in 96% of answers and staff turnover in 83%.
Reddit, Glassdoor, and Indeed supplied 13.6% of citations in those named company checks. These figures describe topics AI raised, not verified problems at the CROs. The study did not validate every claim.
Does being cited by AI mean your CRO is being recommended?
An AI answer can reference your website to explain a clinical topic, name your CRO among potential providers, or recommend your services for a particular requirement. Each tells a different story about your visibility.
Our study illustrates the distinction through Allucent and IQVIA.
Allucent’s website received more citations, while IQVIA appeared in a larger proportion of answers that named the company.
The measures cover different answer sets. They show how often a company was named and how often its website was referenced. They do not establish how frequently a citation resulted in a recommendation.
For CRO marketing teams, the next step is to examine what each appearance contributes to the buyer’s evaluation.
- Source use: Which pages are cited, and what information do they support?
- Provider inclusion: Is your CRO presented as a suitable option for the buyer’s clinical requirement?
- Accuracy: Does the answer describe your capabilities, therapeutic experience and company identity correctly?
For example, an oncology article could help AI explain trial design without the answer recommending its publisher to run the trial.
Reviewing the full answer alongside its sources helps identify what work is needed. A CRO may need stronger evidence of service suitability, clearer descriptions of its capabilities, or corrections to outdated information.
The marketing objective is consistent, accurate inclusion in the clinical buying conversations relevant to your business.
What should CRO marketing leaders change first?
Start with the services and therapeutic areas central to your growth plan. Review the questions relevant pharma and biotech sponsors ask, then examine your coverage across platforms and repeated runs.
Improve the pages supporting those conversations. Connect service descriptions with relevant experience, clear ownership information and current company names.
14% of company naming answers used at least one former company name, usually as historical context.
The website audit also found a structural gap. Among 103 reachable sites, 63% had basic company schema, but only 4% had service schema. None explicitly blocked the checked AI bots in robots.txt.
Service schema is an opportunity to clarify existing information. The audit does not prove that adding it increases recommendations.
FTA’s Search Engineering™ connects work on content, search visibility and citations with the moments when buyers evaluate providers.
For CRO leaders, the priority is to understand which clinical buying questions you already answer credibly and where your public evidence needs work.
FAQs
How should our CRO use FTA’s study to set its marketing priorities?
Start with the buyer roles, services and therapeutic areas relevant to your growth plan. Use the study to identify where competitors appear and which sources support those answers. Prioritise gaps in the clinical buying conversations your firm needs to enter.
Can midsize CROs compete with larger firms in AI recommendations?
FTA’s study found that midsize CROs received 43% of classified provider mentions when the buyer identified as a small biotech, compared with 27% for other buyers. The finding supports making your suitability for smaller pharma and biotech sponsors explicit through relevant experience and delivery evidence.
Can a CRO have a modest overall mention rate but strong visibility in its speciality?
Yes. FTA’s service comparisons showed Cytel leading biostatistics and programming with a Context Authority score of 80, and Certara leading clinical pharmacology and modelling with 78. Assess your position within the services you sell alongside your overall visibility.
What should we check when AI answers questions about our CRO by name?
Review how AI describes your ownership, team continuity, service capabilities and suitability for pharma and biotech sponsors. In FTA’s named CRO checks, ownership appeared in 96% of answers and staff turnover in 83%. Verify the underlying sources before treating those statements as facts.
How does FTA help CROs act on the study’s findings?
FTA’s Search Engineering™ helps connect clinical expertise with the questions pharma and biotech buyers ask. We map relevant queries, strengthen service and therapeutic content, and examine brand mentions and citations. The work starts with the services, buyers and competitive gaps that matter to your CRO.
Do you want more traffic?
