I have spent the last few months living inside CROs RFIs, RFPs, and proposals, and I have come out the other side convinced that most pre-award questionnaires are not designed to support good decisions. They create the illusion of diligence, but they rarely help sponsors answer the only question that matters:
Can this CRO run this trial on time, within budget, and at the quality bar required?
The problem starts with the questions. Most biotech and pharma RFIs are vague, generic, and strangely abstract, even when they look granular. They ask for process descriptions, capabilities, templates, historical experience, system names, headcount, and SOPs. Some of these questions are necessary. Of course a sponsor needs to know whether a CRO has the right systems, compliance infrastructure, privacy policies, and quality processes. But those are table stakes. They tell you whether a CRO is allowed into the room (and let’s be honest, most of them are). They do not tell you whether this CRO is the right partner for this specific trial.
That is the first mistake. The form assesses the CRO as a company, but the trial will be delivered by a specific team.
A sponsor asks, “How many CRAs do you have?” The CRO answers, “182 across the US.” Wrong question, wrong answer. Nobody actually wants the number. They want to know who will be assigned to their study, what else those people are staffed on, what happens when three of them are out sick at once, and who owns the account 18 months later when the person charming them today is gone.
Capability is a corporate fact. Delivery is what these specific people do for this specific protocol when the plan breaks. The form sees the first. The sponsor lives with the second.
This is also why the A-team/B-team problem survives every procurement process. The A-team wins the bid. The B-team runs the trial. This is what I hear from every Biotech companies struggling to hit their milestones.
Yet most questionnaires do not ask the questions that would catch it. Name the project manager and lead CRA you will assign on day one. Send their CVs. Let me interview them, not business development, before award. What else are those exact people staffed on next quarter? Is any of it already slipping? Who is accountable when the bid team has moved on?
The second mistake is that forms confuse artifacts with behavior.
A sponsor asks, “Do you have an SOP for risk-based monitoring?” The CRO says yes. Great. A document exists. That does not mean the organization behaves that way under pressure. It does not tell you whether anyone uses the SOP when a site goes dark, when enrollment is 40 percent behind plan, or when a CRA sees a risk nobody wants to escalate. The better question is: walk us through a real study where your monitoring approach detected a risk before it became a timeline or quality issue. Show us how the issue moved from site observation to escalation to corrective action. Tell us when the process failed and what you changed afterward.
There is a tell here. A CRO can complete 90 percent of these forms by pasting from a reusable answer library it keeps for exactly this purpose. If both sides can automate the exercise, the exercise is not discriminating between vendors. It is paperwork wearing the costume of judgment.
I will concede the one thing worth conceding. Some of these questions are legitimate. Part 11, GDPR, an SOP library, inspection history, disaster recovery: these are pass-fail gates. You need them. But a gate is a qualifier, not a differentiator. The error is not asking them. The error is mistaking the floor for the building, ticking the boxes, and calling the house inspected.
The third mistake is that sponsors ask for success stories when they should be asking for failure forensics.
Most RFIs ask for relevant experience, successful case studies, and references. But success stories tell you what a CRO wants to sell. Failure stories tell you how a CRO learns. Tell me about a similar trial that went badly. What broke? What did it cost? Which assumption did you get wrong? Which early warning signal did you miss? What changed afterward? Give me a reference from a sponsor relationship that was difficult, not just the one that made the brochure.
A CRO that cannot talk honestly about failure either has not learned enough or is not being honest enough.
The fourth mistake is that questionnaires ask for capability, not commitment.
Capability is easy to claim. Commitment costs something. “Can you support recruitment?” is a capability question. “What enrollment rate per site per month will you stand behind, what assumptions does it depend on, and what happens if you miss it?” is a commitment question. “Can you meet this timeline?” is a weak question. “Which part of this timeline is least realistic, and what would need to be true for it to hold?” is a better one.
Same with site feasibility. Sites are asked, “Are you interested in participating in this trial?” But if a site opened the survey and answered it, they are already showing some level of interest. The question tells you almost nothing. A better question is: how would you prioritize this study against the other trials competing for your team’s attention, and why? That tells you whether the site is actually likely to behave like this trial matters.
Same with patient numbers. Asking a site, “How many patients can you enroll?” is not enough. The number is almost meaningless without context. A better question is: how did you arrive at that number? Did you search your EMR? Did you ask the PI? Did you estimate from memory? Did you account for competing trials, screen failure, payer restrictions, referral leakage, and staff capacity? That is the difference between collecting data and assessing confidence.
And then there is the part the form will never do, because it implicates the sponsor.
Half of trial failure is sponsor-caused. Impossible protocols. Frozen budgets. Decisions that take six weeks when the study needs them in six days. The best CROs already know this, and the best ones quietly qualify their sponsors before they sign, because they know a slow, scope-creeping sponsor will sink a perfect operational plan.
So one of the most predictive questions in the entire process is the one the sponsor has to be brave enough to invite: what do you need from us to succeed, and what do sponsors most often fail to give you?
No questionnaire I read had a serious section where the CRO was allowed to push back on the sponsor. Not really. And that is not a small omission. The deepest reason RFIs avoid the questions that predict failure is not just laziness. It is that the honest answers point back at the protocol, the timeline, the budget, and the sponsor’s own decision-making. The RFI is a mirror, and the sponsor does not want to look into it.
The best CROs should not just answer the sponsor’s questions. They should challenge the sponsor’s assumptions. They should be able to say, “This timeline is unrealistic,” “This site mix will not work,” “This endpoint will slow recruitment,” “This budget is missing the operational complexity,” or “This protocol will create avoidable friction at the site level.” A CRO that only says yes is not necessarily a good partner. Sometimes the most valuable CRO is the one willing to tell the sponsor where the plan will break.
The problem is that honest CROs are often punished, while CROs that support delusional timelines are rewarded.
I am a founder myself. I get it. The whole endeavor is high risk. Sometimes the only way to survive the pressure is to carry a level of denial. But it is one thing to deny the possibility of your asset failing so you can keep going. It is another thing to endanger the asset by refusing to engage with operational reality.
So here is the claim, plainly: the questions that predict trial delivery are usually the questions the RFI and RFP refuse to ask.
And notice what these questions have in common. They cannot be answered well on paper. They cannot be answered asynchronously. They cannot be answered by a proposals team copying from an answer library. They surface live, on the actual protocol, with the actual delivery team in the room, and with a sponsor willing to be questioned back.
Which means the fix was never just a better form. A better form still gets you a better-rehearsed answer from the wrong person about the wrong time horizon. The fix is a different instrument.
Put the proposed project manager in a working session on your real protocol. Bring the lead CRA. Bring clinical operations. Bring the people who will actually run the study. Ask them where the plan breaks. Ask them what they would change. Ask them which assumptions they reject. Ask them what they need from you. Score them on judgment, not on whether the SOP exists.
Looking at all these RFIs and RFPs, here is what I see: pharma and biotech pre-award due diligence often feels like a hunt for data to assess CRO capabilities. But capability is not delivery. Capability tells you what the CRO could do in principle. Delivery tells you what this team will actually do on this protocol, under real constraints, when the plan starts to break.
Sponsors do not need longer questionnaires. They need better instruments for judgment. They need to know which CRO understands the study, which team will run it, which risks they see, which assumptions they reject, which numbers deserve confidence, and which commitments they are willing to stand behind.
As a sponsor, what you are buying is not clinical trial execution as a service. That is the how. What you are buying is a successful trial operation. What you want is to get to an answer: should I move forward with this asset, yes or no?
What you are buying is a CRO that can help you ship that answer on time and on budget.
Originally published in Trials & Triumphs, Yendou’s newsletter on the work between RFI and First Patient In.
Subscribe for fresh ideas, straight to your inbox →