How to Find Your CRO Partner

The CRO proposal is the number in the budget. Site activation is the number that sets the timeline.

The proposal total is what reaches the budget committee. Site activation, enrollment and team continuity are what determine whether the study finishes on the date in that proposal. The two are not the same number, and the gap between them is where most of the cost of a clinical program is actually decided.

The outsourcing question sponsors search for is usually “CRO or CDMO?” That question has a short answer: a CRO runs the study, a CDMO makes the product. The harder question, and the one this briefing addresses, is which CRO, and how to choose one on evidence rather than on a presentation.

What the Data Shows

Findings from Tufts Center for the Study of Drug Development (Tufts CSDD) research, US federal regulation and ICH guidance, and CDMO Hub analysis.

• A day of delay costs roughly $500,000 to $800,000 in lost sales. The peer-reviewed paper reports approximately $500,000; the Tufts white paper on the same study reports approximately $800,000. Both replace the $4–5 million figure that has circulated since 1993.
• Running the trial itself costs about $40,000 a day across Phase II and III, and $55,716 a day in Phase III.
• Study start-up is getting slower, not faster. Over the past decade, the time from protocol approval to first patient, first visit rose 44% in Phase II and 27% in Phase III (Tufts CSDD, 2025).
• Almost half of sites miss their enrollment target. In a Tufts CSDD analysis of nearly 16,000 sites, 11% enrolled no one and 37% under-enrolled. The data are from 2008–2010 and remain the most cited site-level benchmark.
• More than 75% of protocols need at least one substantial amendment. The median direct cost of implementing one was $141,000 in Phase II and $535,000 in Phase III (Tufts CSDD).
• Oversight cannot be outsourced. A sponsor can transfer trial obligations to a CRO in writing, but ICH E6(R3) keeps ultimate responsibility for trial quality and data integrity with the sponsor.
• CRO proposals are not directly comparable (CDMO Hub analysis). Each is priced on its own assumptions for site count, monitoring model and pass-through costs, so the headline totals measure different studies.

The CRO Landscape, Mapped by Fit

The first sourcing decision is not which CRO but which kind. The models below are not a ranking. Each fits a different program, and each has a characteristic way of going wrong.

The Point Most Often Missed

The team at the bid-defense meeting is not necessarily the team that runs your study. Proposals are presented by senior staff; studies are staffed from whoever is available at kickoff, and the two groups can differ. Team continuity can be contracted. Name the key personnel in the work order, give yourself approval over any replacement, and set a notice period. Without that, continuity is a promise, not a term.

Where CRO Selection Goes Wrong

Most selection failures are set in motion before the first proposal arrives. The process has seven stages, and each has a characteristic failure.

1. Requirement definition. The sponsor issues an RFP before settling its own country strategy, patient population assumptions or monitoring approach. Every CRO then fills the gaps with its own assumptions, and the proposals stop being comparable. Fix the assumptions that drive cost before going out.
2. Longlist. Built from names the team already knows, conference booths and personal networks. The result is a list of the most visible CROs, not the best-fitted ones. Start from the program’s requirements: phase, therapeutic area, countries, modality and patient population.
3. RFI. Free-text questions produce marketing answers. Ask for specific, dated operating data in a fixed format, so that answers can be placed side by side.
4. RFP. Each CRO builds its budget on different site counts, visit schedules, monitoring intensity and pass-through handling. Issue a fixed budget grid with the key assumptions pre-filled, and ask for deviations to be listed separately.
5. Bid defense. The meeting rewards presentation skill. Require the proposed project manager and lead clinical research associate (CRA) to attend, and ask them to walk through their last comparable study.
6. Contracting. The master services agreement and work order are negotiated for price. The terms that protect the timeline (key personnel, performance metrics, change-order rules, data ownership) are left to standard wording.
7. Kickoff. No oversight plan is agreed. The sponsor learns about problems from monthly reports rather than from agreed metrics with thresholds and escalation paths.

A worked illustration (hypothetical). Two proposals arrive for the same Phase II study: CRO A at $9.2 million, CRO B at $11.0 million. CRO A assumed 40 sites, reduced on-site monitoring and investigator fees billed as pass-through outside the total. CRO B assumed 55 sites, a higher monitoring frequency and pass-through costs inside the total. Normalized to the same site count, monitoring model and cost boundary, the gap can close or reverse. The lower total was a different study, not a cheaper one.

The Rulebook, and What It Lets You Ask For

The regulations that govern CRO use do not tell sponsors which CRO to choose. They define what the sponsor remains responsible for, which is what gives the sponsor the right to ask for evidence.

The Point Most Often Missed

These rules add conditions rather than prohibitions. None of them stops a sponsor from outsourcing a whole trial. All of them keep the sponsor accountable for what it outsources, which means sponsor oversight has to be designed into the CRO relationship from the RFP stage rather than added after kickoff.

How to Score a CRO: A Scorecard

Every CRO will describe itself as experienced, responsive and quality-focused. The scorecard below replaces descriptions with evidence. For each criterion it separates what a CRO can claim from what it can document, and tags the lever the criterion moves: TIME, RISK or RETURN.

Weighting depends on the program. A first-in-human study with a small team weights continuity and quality highest. A multi-country Phase III weights site network and budget structure highest. Agree the weights before proposals arrive, so they cannot be adjusted to fit a preferred bidder.

The Constraint Nobody Prices In: Once Signed, You Are Locked In

Switching CRO mid-study is possible, but rarely economic. It means migrating or rebuilding the database, re-papering site contracts, retraining a new team on the protocol and re-establishing site relationships, while the study keeps running. At roughly $40,000 a day in direct trial costs, and far more in lost sales for a late-stage program, the switching cost usually exceeds any saving in the original bid.

The consequence is that a sponsor’s leverage over its CRO is highest before signature and falls sharply after it. The protections that matter have to be negotiated while that leverage still exists:

• Key-personnel clauses: named project manager and lead CRA, sponsor approval of replacements, and a notice period.
• Performance metrics with consequences: site activation, enrollment and data-entry timeliness against agreed thresholds, linked to escalation or fees.
• Change-order rules: what counts as out of scope, the rate card for changes, and approval thresholds before work starts.
• Data ownership and exit: export format, timing and cost of data transfer at contract end, whether the study ends or the CRO is replaced.

Any CRO claim of faster trials should be tested against one question: which studies, in which countries, measured from what baseline, and can you show us the records?

Diligence Questions That Produce a Real Answer

• Site activation. What was your median time from site selection to site activation in [our countries] over the last 24 months, taken from your own records rather than industry benchmarks?
• Enrollment. On your last three studies in this indication, what share of activated sites enrolled zero patients, and what share met target?
• Team. What was project manager and CRA turnover on those studies? Which named people will run ours, and will you contract them?
• Change orders. On comparable studies, what was total change-order value as a share of the original contract, and what were the three most common causes?
• Inspections. What were the outcomes of your most recent regulatory inspections, and how were the findings closed?
• Oversight. What metrics and dashboards will we see, how often, and with what thresholds for escalation?
• Data exit. At contract end, in what format and on what timeline do we receive the complete dataset and trial master file, and at what cost?
• Supply interface. Name a study where the sponsor used a third-party CDMO or clinical supply provider. Who owned the enrollment forecast that drove drug supply?
• Transfer of obligations. Which sponsor obligations do you propose to assume in writing, and which remain with us?

Why This Matters for Sponsors

1. Headline price is not cost. Proposals are built on different assumptions. Normalize site count, monitoring model and pass-through costs before comparing, or the comparison is between different studies.
2. Site activation data predicts the timeline; brand does not. Start-up durations have lengthened over the past decade, and almost half of sites miss their enrollment targets. A CRO’s own activation and enrollment record in your countries is the most direct evidence of when your study will finish.
3. The contract is the only protection that survives signature. Key personnel, performance metrics, change-order rules and data exit terms have to be negotiated before signing, because switching later is rarely economic.
4. Oversight stays with the sponsor. Under ICH E6(R3) and 21 CFR 312.52, responsibility does not transfer with the work. Choose a CRO whose systems give you the visibility to exercise it.

The CDMO Hub View

Three sourcing strategies are in use, and they are not interchangeable. Large full-service CROs offer global reach and a single accountable provider; they fit multi-country pivotal studies, but a smaller sponsor’s study competes for attention with larger clients. Mid-size and specialist CROs offer senior attention and investigator relationships in a defined therapeutic area or region; they fit early-phase and focused programs, but their capacity in each proposed country needs checking. FSP and functional mixes give the sponsor the most control; they fit sponsors with internal clinical operations, but the sponsor then owns every hand-off between providers.

Our expectation is that price and therapeutic experience will remain the primary selection criteria within this planning cycle. What will change first is how they are tested. Sponsors that score CROs on dated, CRO-specific operating data (activation times, enrollment by site, team turnover, change-order history) will reach first patient in sooner than sponsors that score on proposals, because they will have selected against the causes of delay rather than discovering them after kickoff.

A Note on Scope

This briefing covers the selection of a CRO for clinical trial services. It does not assess any individual CRO’s quality, capacity, compliance history or commercial terms, and no company is named or recommended. Benchmark figures are drawn from published Tufts CSDD research and describe industry averages, not the performance of any single provider. The site-enrollment benchmark uses 2008–2010 data and is cited as the most widely used published figure, not as a current measurement. The worked budget illustration is hypothetical. The conclusion that verified operating data predicts delivery better than proposals is CDMO Hub’s analytical view.

CDMO Hub Insight

The next improvement in trial timelines will come less from new CRO capabilities than from how sponsors choose between existing ones. Site activation, enrollment by site and team continuity are measurable before a contract is signed, and they explain more of the timeline than anything in a proposal.

For outsourcing teams, one request separates a real capability from a presented one: show us your median site activation time in our countries, from your own records. A CRO that can answer has measured its own performance. One that cannot has submitted a proposal.

Looking for a CRO partner?

Describe your study once in SmartRFI and send it to a shortlist of CROs. Responses come back in a comparable format, ready for a decision you can defend internally. Start at CDMOHub.ai.

Sources

• Cost of delay: Smith Z, DiMasi J, Getz K. “New Estimates on the Cost of a Delay Day in Drug Development.” Therapeutic Innovation & Regulatory Science, 2024; Tufts CSDD, Quantifying the Value of a Day of Delay in Drug Development, white paper, August 2024.
• Study start-up durations and amendment prevalence: Tufts CSDD Impact Report, Vol. 27 No. 6 (November/December 2025) and Vol. 25 No. 2 (March/April 2023), as presented in WCG Clinical protocol amendment webinar, 2026.
• Amendment cost: Tufts CSDD, press release on substantial protocol amendments, 14 January 2016.
• Site enrollment performance: Getz K, “Enrollment Performance: Weighing the ‘Facts’”, Applied Clinical Trials, 2012; Tufts CSDD press release, January 2013.
• ICH E6(R3): ICH Harmonised Guideline for Good Clinical Practice E6(R3), adopted 6 January 2025; EU effective date per Sidley Austin summary, February 2025.
• Transfer of obligations: 21 CFR 312.52.
• EU Clinical Trials Regulation: Regulation (EU) No 536/2014.

Figures and characterizations are current as of September 2026. Corrections are welcomed and will be carried in the following issue. CDMO Hub · CDMOHub.ai