Building a Competitive Research Team for Large Grant Applications

Grant reviewers rarely reject a proposal because the science is boring. Far more often, they reject it because they cannot tell whether the people behind it can actually pull the work off. Building a competitive research team for a large grant application means proving, on paper, that your principal investigator, co-investigators, biostatistician, and coordinator each bring something the project genuinely needs, and that they have already worked together well enough to trust the plan. NIH changed how it scores that proof in 2025, and it changed the paperwork behind letters of collaboration and biosketches again in 2026. A team description written for the old rules can now cost you points under the new ones.

What Makes a Research Team Competitive for Large Grants

Funders do not score team size. A twelve-person roster with overlapping, vague roles reads weaker than four people whose expertise maps cleanly onto the aims. Reviewers are trained to ask a narrow question here, does this specific group of people have what it takes to do this specific project, not whether the roster looks impressive on paper.

Interviews with members of research teams that lasted and teams that fell apart point to the same handful of ingredients. Bennett and Gadlin, in a study of thriving and struggling NIH-connected research teams published in the Journal of Investigative Medicine, found that durable, high-performing teams share trust, a genuinely shared vision, and deliberate early effort spent choosing and building the team rather than assembling it at the last minute.¹ None of that shows up automatically in a biosketch. Reviewers only see it when it is written into the team description, the letters of collaboration, and the leadership plan in specific, checkable language.

Two questions are worth answering honestly before drafting a single sentence about your team. Does every named investigator have a task in the Research Strategy that only they are qualified to do? Has the group produced anything together before, even a pilot dataset or a joint conference abstract, that reviewers can point to as evidence the collaboration already works? For a closer look at the specific tradeoffs of working across fields, our guide to building interdisciplinary research teams covers how mismatched disciplinary cultures can undercut an otherwise strong roster.

The Core Roles a Grant Reviewer Expects to See

A large grant application typically needs more roles filled than a single lab can cover alone. Reviewers look for each of the following, not because a checklist demands it, but because a project of real size usually stalls without them.

• Principal Investigator (PI) or contact PI. Owns the scientific vision and, on multi-PI awards, coordinates the leadership team and communicates with the funding agency.
• Co-Investigators. Bring a specific method, population, or technical skill the PI does not have, tied directly to one or more aims.
• Biostatistician. Designs the power calculation and analysis plan, ideally involved since the aims were drafted rather than added afterward to check the math. A missing or late-added statistician is one of the fastest ways to trigger a rigor and feasibility concern. Before that conversation, it is worth sanity-checking your own numbers with a free sample size calculator, and if the project involves patient-reported or other complex outcomes, our guide to statistical methods for patient-reported outcomes is a useful starting point.
• Project or research coordinator. Manages recruitment, regulatory submissions, data collection logistics, and the day-to-day schedule that keeps a multi-year award on track. Regulatory timelines matter here too; see our notes on cutting IRB review time for how coordinators typically shave months off this step.
• Key personnel and consultants. Named individuals, often at partner institutions, who provide a specific service, technology, or population access documented through a letter of collaboration.
• Trainees and research assistants. Support data collection and management, and on career development awards, are the people the PI is proposing to train.

Every one of these roles needs a named person attached to it by the time the application goes in, not a placeholder like “TBD” or “to be recruited.” Reviewers read an unfilled role as a project that is not actually ready to start.

How NIH Scores Your Team Under the Simplified Peer Review Framework

NIH stopped scoring Investigator and Environment as their own numbered criteria for applications with due dates on or after January 25, 2025. The old five-criterion system, Significance, Innovation, Approach, Investigator, and Environment, folded into three factors under what NIH calls the Simplified Review Framework.²

Factor 1, Importance of the Research, combines Significance and Innovation and gets a numeric score from 1 to 9. Factor 2, Rigor and Feasibility, is essentially the old Approach criterion and also gets a numeric score. Factor 3, Expertise and Resources, folds together what used to be Investigator and Environment, and it works differently. Reviewers do not give it a number. They judge it as sufficient or not sufficient for the specific project proposed, and if they see a gap, they have to explain it in writing rather than quietly docking a point.²

That binary judgment matters more for team-building than a numeric score ever did. A seven out of nine on the old Investigator criterion was ambiguous, good enough to average out against a strong Approach score. A written statement that your team’s expertise and resources are “not sufficient” for a specific aim is not something an otherwise strong Significance section can absorb. Multi-PI applications get an added layer of scrutiny here, since reviewers also weigh how well the leadership team is coordinated, not just what each PI individually brings.²

Single PI, Co-Investigators, or Multiple PI, Choosing Your Leadership Structure

NIH created the multiple principal investigator option in 2006 after its Bioengineering Consortium and the Office of Science and Technology Policy both pushed for a structure that fit team-based science better than a single hierarchy.³ The option supplements, rather than replaces, the traditional single-PI model, and choosing wrong in either direction can weaken an application.

Multi-PI makes sense when two or more people are genuinely leading the science together, particularly across departments or institutions, and no single person should hold sole authority over the direction of the work. Single PI with co-investigators still fits better when one person drives the project and others contribute defined, specialized pieces of it.

Every multi-PI application needs a Leadership Plan, and NIH is specific about what belongs in it. It has to spell out each PI’s role, how responsibilities are divided, and how the leadership team will actually work together day to day, not just in principle.³ A Leadership Plan that reads as boilerplate, swapping in different PI names but otherwise identical to a template, tends to draw exactly the kind of written concern the Expertise and Resources factor now allows reviewers to raise. Our guide to managing multi-PI proposals walks through how to write a Leadership Plan reviewers find credible.

Biosketches, Other Support, and the 2026 SciENcv Requirement

Every named investigator’s paperwork changed again this year. As of January 25, 2026, NIH requires the new Common Forms for the Biographical Sketch and for Current and Pending (Other) Support on all applications, and on Just-in-Time, RPPR, and Prior Approval submissions filed on or after that date.⁴ The old NIH-specific biosketch format is being retired in favor of forms shared across other federal agencies.

SciENcv is no longer optional. Each senior and key person on your team has to link an ORCID identifier to their eRA Commons profile, build a biosketch and Other Support disclosure inside SciENcv, and personally certify the resulting PDF, including certifying non-participation in Malign Foreign Talent Recruitment Programs. Administrators can help populate the data, but they cannot certify on someone else’s behalf.⁴ NIH is phasing in enforcement as system warnings first, moving to submission-blocking errors by February 6, 2026.

For a team assembled from multiple institutions, this is a coordination problem as much as a paperwork one. Build in time for every co-investigator, not just the PI, to set up SciENcv and certify their own documents well before the internal deadline, since this step cannot be delegated the way earlier biosketch drafts often were.

Letters of Collaboration Have Replaced Letters of Support

NIH is retiring the traditional letter of support entirely, moving to a shorter letter of collaboration that lines up with how NSF has long handled the same document.⁵ The change takes effect with Forms J in fiscal year 2027, and applications submitted after that point that still include old-style letters of support risk being returned without review.

NIH designed the new letter of collaboration to be narrow on purpose. It is capped at 100 words, and it can only document collaborative intent or a specific material contribution, not praise for the applicant or an opinion on the project’s merit. Letters cannot come from trainees, graduate students, postdocs, or technicians working in the PI’s own lab, closing off a workaround some applications used to lean on. NIH even offers a model sentence, stating an intent to collaborate and commit resources as detailed in the application, if selected for funding.⁵

For a team-heavy application, this is good news disguised as a formatting rule. A padded, effusive support letter never actually proved a collaborator would show up and do the work. A one-hundred-word letter that names a specific resource or task is harder to pad and easier for a reviewer to check against what the Research Strategy claims that person will contribute.

Building a Competitive Research Team for NSF Proposals

NSF reviews every proposal against two criteria, Intellectual Merit and Broader Impacts, and both criteria fold team qualifications into the same rating rather than scoring the investigators separately the way NIH used to.⁶ Reviewers are explicitly asked how well qualified the individual, team, or organization is to carry out the proposed work, and whether adequate resources are available, either inside the team or through outside collaborations, to get it done.⁶

That framing rewards a team that can show, rather than just claim, capability. A proposal listing collaborators without describing what each one specifically adds to Intellectual Merit or Broader Impacts reads weaker than one tying every named person to a task under one of the two criteria. Broader Impacts in particular tends to reward teams that already include a partner organization, an evaluator, or a community collaborator by the time of submission, rather than promising to find one later.

Multi-institution NSF teams face the same practical problem NIH teams do, coordinating people who answer to different departments and different indirect cost rates. Our guide to writing an NSF CAREER proposal covers how a single early-career PI builds a credible team and mentoring structure around a CAREER award specifically.

Where Reviewers Most Often Find Weaknesses in the Research Team

NIH’s Center for Scientific Review has long circulated a well-known list of the problems reviewers cite most often across applications, and several map directly onto team composition even though the list predates the current framework.⁷ “Lack of experience in the essential methodology” is one of the most common, and it almost always traces back to a missing role rather than a missing idea, a proposal built around a statistical method nobody on the team has published using, or an imaging technique described in detail with no co-investigator credited for running it.

A few patterns show up repeatedly in team sections specifically, separate from CSR’s general list.

• A biostatistician named in the budget but never mentioned in the Approach section, leaving their actual involvement in the analysis plan unclear.
• Co-investigators with no evidence they have worked together before, no joint publication, no shared pilot data, nothing beyond a signature on a letter.
• Effort levels that do not match the role described, such as a co-investigator committing one percent effort to a task the Research Strategy calls central to an aim.

None of these are fatal on their own, but each is exactly the kind of concern the Expertise and Resources factor was built to surface in writing rather than bury inside an average score. Separate research on NIH peer review found substantial disagreement between reviewers scoring the very same applications, which is its own argument for writing an unambiguous team section.⁸ The less room reviewers have to guess who is doing what, the less your score depends on which reviewer happened to read it. If a summary statement has already flagged your team, our guide to addressing reviewer comments in a resubmission walks through how to respond point by point.

A Realistic Timeline for Assembling Your Team Before You Write

Team assembly has to finish well before the writing does, and NIAID’s own planning guidance gives a useful floor for how long that takes. For a straightforward R01 without vertebrate animals or human subjects, NIAID recommends at least two months of dedicated preparation time, another two months to write, roughly two weeks for colleagues to give feedback, and a final week to check and recheck the application before it goes to your institution’s business office.⁹ That is four and a half months at a minimum, and multi-PI or multi-institution teams should plan for longer.

Preparation time is where team-building actually happens, not writing time. Co-investigators need to be approached, roles negotiated, and letters of collaboration requested early enough that a collaborator is not asked to sign something the week of the deadline. SciENcv certification, discussed above, adds a step that used to take an afternoon and now takes coordination across every named investigator. Most sizeable teams also owe the funder a joint Data Management Plan by submission; our guide to data management plan requirements covers what NIH and NSF now expect there.

A rough working schedule for the team-building side alone looks like this.

• Four to five months out. Identify gaps in your own expertise and start approaching co-investigators and a biostatistician.
• Three months out. Confirm roles, effort levels, and who is requesting which letters of collaboration.
• Two months out. Everyone drafts their sections and certifies SciENcv documents in parallel with the writing.
• Two weeks out. Circulate the full draft to co-investigators for a final check on their own sections.
• One week out. Final proofread and submission through your institution’s business office, which often needs five to ten business days on its own.

Before and After, Rewriting a Weak Team Paragraph Into a Strong One

The difference between a weak and a strong team section rarely comes down to the underlying people. It usually comes down to whether the paragraph gives a reviewer anything specific to evaluate. The two versions below are illustrative examples written to show the pattern, not text pulled from an actual funded application.

Weak version

“Our research team brings together extensive experience across multiple disciplines relevant to this project. Dr. Smith has a strong publication record in the field, and Dr. Lee will provide statistical support. Both investigators are highly qualified and have the skills necessary to ensure this project’s success. The team is well-positioned to carry out the proposed work and looks forward to a productive collaboration.”

Strong version

“Dr. Smith (PI, 30% effort) has led three funded studies using the proposed single-cell sequencing protocol and will direct Aims 1 and 2. Dr. Lee (biostatistician, 15% effort) designed the power calculation in Section C and has co-authored two prior papers with Dr. Smith using the longitudinal mixed-effects approach proposed in Aim 3. Dr. Lee joins the study at the design stage, not after data collection, and holds primary responsibility for the interim analysis described in the Data Safety and Monitoring Plan. A letter of collaboration from the Regional Biorepository (Appendix B) confirms sample access for Aim 2.”

What changed is specific and repeatable.

• Named tasks replaced adjectives. “Highly qualified” became a specific protocol, a specific aim, and a specific number of prior funded studies.
• Effort percentages appeared next to each name, so the budget and the narrative agree with each other.
• Prior collaboration got named evidence, two co-authored papers using the same method, rather than an assertion that the team works well together.
• The biostatistician’s role got tied to a specific section of the application, not left as a general promise of “statistical support.”
• A letter of collaboration backed a specific resource claim instead of a general statement of enthusiasm.

Every one of those changes answers a question the Expertise and Resources factor now asks explicitly, leaving a reviewer very little to guess at.

What the Evidence Says About Why Research Teams Succeed

Two studies, read together, describe a fairly consistent picture of what separates research teams that hold together from ones that don’t. Bennett and Gadlin’s interviews with members of both thriving and struggling NIH-connected teams point to trust as the central ingredient, alongside a shared vision the whole team actually believes in and deliberate effort spent choosing team members rather than defaulting to whoever happened to be available.¹

Brown and colleagues, writing about building an interdisciplinary cardio-oncology research team, arrived at a more structured version of the same idea. Their paper lays out eight principles, team building, leadership, structured discussions, mentorship, training, planning, communication, and collaboration, and argues most of them are never formally taught in health professional training, so teams either stumble into them or don’t.¹⁰

Neither paper describes a shortcut. Both point to the same unglamorous work, deciding early who belongs on the team and why, then protecting the relationships that make the collaboration real. That happens to be exactly what a reviewer is trying to verify when reading your Expertise and Resources section.

How Editverse Strengthens Your Competitive Research Team

A strong team in practice still loses points on paper when the application undersells it. Editverse works on that specific problem for research teams applying to NIH, NSF, and other major funders.

For the rest of the application beyond the team section itself, our companion guide on how to write an NIH grant proposal covers the Specific Aims page, the Research Strategy, and the statistical plan in depth.

Frequently Asked Questions

How many people does a competitive research team need for a large grant?

There is no fixed number. Reviewers care whether every role required by the Research Strategy is filled by a named, qualified person, not whether the roster hits a particular size. A four-person team with no gaps beats a ten-person team with an unfilled statistical role.

Do I need a named biostatistician even for a small project?

If the application includes any quantitative analysis plan, a named biostatistician involved early enough to shape the power calculation is safer than adding one later to check the numbers. “Lack of experience in the essential methodology” remains one of the most common problems reviewers cite in applications.⁷

What is a Leadership Plan and when do I need one?

NIH requires a Leadership Plan for every multiple-PI application. It has to describe each PI’s role, how responsibilities are divided, and how the leadership team will work together, and it needs to read as specific to your project rather than boilerplate.³

What replaced NIH letters of support?

Letters of collaboration, capped at 100 words, phasing in with Forms J in fiscal year 2027. They can document collaborative intent or a specific contribution only, not an endorsement, and they cannot come from trainees, students, postdocs, or technicians in the PI’s own lab.⁵

Does my whole team need to set up SciENcv, or just the PI?

Every senior and key person on the application needs an ORCID linked to eRA Commons and a certified SciENcv biosketch and Other Support form, effective for submissions on or after January 25, 2026. This step cannot be delegated to an administrator.⁴

What if my institution does not have every expert my project needs?

Bring in a named collaborator at another institution, backed by a letter of collaboration describing their specific contribution. Reviewers evaluate whether adequate resources are available “either inside the team or through outside collaborations” under both the NIH and NSF frameworks, so an external expert with a clear, documented role is an acceptable answer to a gap.^2,6

References

1. Bennett LM, Gadlin H. Collaboration and team science, from theory to practice. J Investig Med. 2013;61(5):768-775. doi: 10.2310/JIM.0b013e318250871d. PMID: 22525233
2. National Institutes of Health. Simplified Review Framework for NIH Research Project Grant Applications. Grants & Funding. grants.nih.gov
3. National Institutes of Health. Multiple Principal Investigators. Grants & Funding. grants.nih.gov
4. National Institutes of Health. NOT-OD-26-018. Implementation of Common Forms for Biographical Sketch and Current and Pending (Other) Support for Due Dates on or after January 25, 2026. grants.nih.gov
5. National Institutes of Health. NOT-OD-26-094. NIH Is Replacing Letters of Support with Letters of Collaboration to Reduce Administrative Burden. grants.nih.gov
6. National Science Foundation. Merit Review Criteria. How We Make Funding Decisions. nsf.gov
7. National Institutes of Health, Center for Scientific Review. Top 10 Problems Reviewers Cite in Applications. massgeneral.org (NIH CSR)
8. Pier EL, Brauer M, Filut A, Kaatz A, Raclaw J, et al. Low agreement among reviewers evaluating the same NIH grant applications. Proc Natl Acad Sci U S A. 2018;115(12):2952-2957. doi: 10.1073/pnas.1714379115. PMID: 29507248
9. National Institute of Allergy and Infectious Diseases. Timeline to Plan and Write Your Application. niaid.nih.gov
10. Brown SA, et al. Team principles for successful interdisciplinary research teams. Am Heart J Plus. 2023. PMC10946054