July 29, 2026 at 12:00 PM ET

Webinar Recap: Crafting Grant Proposals in the Life Sciences

Writing NIH Proposals that Build Reviewer Confidence.

Two scientifically strong proposals can be received very differently. Dr. Brandi Matson, a neuroscientist and founder of Life Science Editors, has spent about 15 years working with scientists on how they communicate complex work, and she built this session around a single question: why does good science alone so rarely guarantee a good score?

Her answer is that investigators think they are writing for NIH, but they are really writing for a specific group of up to 20 people who have to understand the science, evaluate it, discuss it, and assign it a score. As paylines tighten, the margin for reviewer uncertainty tightens with them. Grantsmanship, she argued, is not a soft skill. It is part of institutional strategy, and research administrators sit at the front line of it.

A funding landscape with no room for uncertainty

The numbers set the stakes. Success rates for R01-equivalent grants have fallen from roughly 23% in 2023 to about 13% in 2025. At the National Cancer Institute, a fundable score has dropped to the 4th percentile, and NCI funded about 400 grants in 2025 rather than the expected 700.

The takeaway is not despair, it is precision. Competitive applications are the ones that reduce reviewer effort and increase reviewer confidence. As competition gets tighter, the tolerance for anything a reviewer cannot quickly understand or defend gets smaller.

Research administrators are the invisible architects Administrators are the last critical reader before submission, and they see far more proposals than any single faculty member will in a career. That vantage point lets them catch patterns a PI can miss.

The job is not to solve the science. It is to notice what Dr. Matson calls reviewer friction before a reviewer ever does:

• Missing information
• Unclear logic
• Inconsistent messaging
• Rushed revisions
• Compliance gaps Her framing for the whole role: your most valuable contribution is recognizing that the problem exists before the reviewer does.

If you have read this three times and still are not sure what the central hypothesis is, odds are a tired reviewer will not be either.

The new NIH framework, in plain terms

NIH simplified its scoring into three factors, and Dr. Matson offered a working translation of each:

• Factor 1: Importance (scored). Merges significance and innovation. Does this address an unmet need and meaningfully advance the field? This factor sets the ceiling for your impact score. It creates reviewer enthusiasm.
• Factor 2: Rigor and Feasibility (scored). The old approach section. Are the methods rigorous and reproducible, and is the study realistically executable? This is the operational center of the grant and usually where the study section discussion lands. It drives reviewer confidence.
• Factor 3: Expertise and Resources (not scored). Pass or fail. Is this the right team, with adequate resources, to answer these questions? It is not scored, but it can quietly drive the scores for Factors 1 and 2. The simplified framework did not lower expectations. The question underneath every factor is the same: do I believe this project can succeed? Feasibility has to be demonstrated, not implied.

The specific aims page is the proposal

If there was one idea to carry home, this was it. During study section, the people who vote on your score may never have read every page. They follow the discussion by looking at the one page in front of them, and that page is the specific aims page. It becomes the lens through which they interpret everything they hear.

If the aims page is confusing, the reviewer spends the research strategy trying to reconstruct your argument. If it is clear, the strategy simply supplies evidence for an argument they already understand, and the reviewer becomes your advocate.

The simplest internal test: take the aims page away. From the rest of the application alone, can you explain the project to a colleague? If not, keep working on it. For extra rigor, read it first thing in the morning, under-caffeinated, and see if it still holds.

Build Factor 1 as one continuous argument

Significance and innovation are not independent claims. Significance answers "does this problem matter?" Innovation answers "does this approach move us forward?" Lead with innovation before establishing the gap, and it reads as incremental. Establish the conceptual gap first, and innovation becomes the answer to it.

Treat Factor 1 as a single line of reasoning that runs from unmet need, to scientific advance, to impact. If this succeeds, what changes?

Language that builds confidence

Reviewers are not editing your prose. They are reading scientific confidence in the words you choose, and they interpret vague language as uncertainty, specifically the uncertainty that you cannot do what you propose. The fix is precision, specificity, evidence, and falsifiability.

Low-confidence: "We will investigate whether X affects Y."

• Reviewer-confident: "We hypothesize that X activates Y via Z, measurable by [assay] in [model] within [timeframe]."

Low-confidence: "Our team has extensive experience."

• Reviewer-confident: Show who has the experience, where they published, and what they have already demonstrated. Don't tell me you're experienced, show me.

She walked an Alzheimer's example from standard prose to a defensible argument. "One of the most important and devastating diseases" became "6.7 million Americans live with Alzheimer's disease, yet no disease-modifying therapy targets inflammation." "A lot of research has shown" became a cited meta-analysis of postmortem studies demonstrating elevated IL-1 beta, TNF-alpha, and complement C1q in the hippocampus, followed by the real open question: whether that activation precedes or follows synaptic loss, and why the answer changes the therapeutic strategy.

This is not better, bigger adjectives. This is better evidence.

The same discipline applies to population and inclusion language. Rather than stating a commitment to diversity in the abstract, name the biologic and analytical groups being studied and tie the inclusion strategy to the study rationale, for example that excluding female participants from prior trials lowers the predictive validity of current risk models for women. Reviewers evaluate what you demonstrate, not what you intend.

Inside the room: how a score is formed

Assigned reviewers score independently, then discuss. A primary reviewer reads the whole proposal and leads the critique, a secondary and a reader add perspective, and the rest of the panel weighs in on strength, rigor, feasibility, and importance, if the application is not triaged first.

Your primary reviewer is your advocate. The goal of every sentence is to let that person articulate the problem, explain the innovation, defend the approach, and, most importantly, answer another panelist who raises a feasibility concern. Consensus is shaped by clarity, confidence, and defensibility during that discussion.

A useful piece of homework: you can look up your assigned study section's roster online. Counting the PhDs and MDs, and noting whether members lean mechanistic or clinical, tells you how to tailor the same science for the room that will read it. A clinically oriented panel wants patient burden and translational impact plus the right power analysis, a basic-science panel wants mechanistic depth and conceptual advance.

Confidence compounds, and so does uncertainty One unanswered question tends to create another. A missing control or an unclear feasibility point can change how a reviewer reads every sentence that follows, and many applications are triaged before discussion, so early impressions matter.

Confidence compounds the same way. A strong aims page sets the frame, a confident primary reviewer explains the science clearly, and a realistic experimental plan gives the room something it can comfortably defend. The common trap is assuming ambition signals excellence. In peer review, ambition without feasibility reads as risk.

Reviewers are more comfortable funding achievable excellence than ambitious uncertainty.

More aims, more experiments, more technologies, and more endpoints do not automatically improve a proposal. The only question that matters is whether each element makes the scientific argument stronger and more believable.

Questions from the room

• Conflicting reviewer critiques. When two critiques seem to pull against each other, map where each sits in the grant, then have the PI contact the program officer for clarification. Often the conflict is two readings of the same vague phrase, and the PO can tell you which concern weighed more.

• Going in blind on study section. PIs can request an assignment in the cover letter and on the submission forms, but assignments are not guaranteed. The most reliable move is to contact the program officer handling the funding announcement before submitting. As Dr. Matson put it, the program officer is your best friend right now.

• Specificity versus clarity. The specific aims page is the abstract of the entire proposal. Put the most relevant details on page one so a reviewer can understand it there, then elaborate in the body. Repeat the same key phrases from the aims page in the proposal so the reviewer connects the two.

• How much preliminary data. She offered a rough split for the aims page: about half for background and preliminary data, roughly a quarter for the aims themselves with hypotheses and expected findings, and the rest for impact. You do not need your whole portfolio, you need the "light bulb" experiment that sent you down this path, and if you are challenging dogma, the counterpoints.

• Disclosing AI use. The guidance is evolving. As of the session, NIH did not want AI used in proposal preparation. Tightening sentence-level language or using a tool as a sparring partner may help, figures are less clear, and the honest advice is to keep a running tab on current NIH policy.

• Research that does not fit NIH priorities. Reframe the work for funder priorities rather than abandoning it, and do not box yourself into NIH. Consider DOD, DARPA, the Department of Energy, and foundation grants to keep the core of your research portfolio active.

Key takeaways

NIH rewards rigor, feasibility, and clarity, and feasibility has to be demonstrated rather than implied. The specific aims page is the proposal. Reviewers score what they can quickly understand and defend. Vague language reads as uncertainty. Precision, specificity, evidence, and falsifiability build confidence. Your primary reviewer is your advocate. Write so they can explain and defend your science to the room. Grantsmanship is institutional strategy, and research administrators can influence a score before the application ever reaches study section. "Every proposal teaches reviewers what to believe about the science. The question is whether it teaches them 'this is interesting, but I don't know if they can do it,' or 'this is important, the logic is sound, this team can execute it, and I can defend this score.'"

About the speaker

Dr. Brandi Matson is a neuroscientist by training whose career spans neuroscience, oncology, drug development, translational medicine, research development, and scientific publishing, including time as an editor at Neuron.

She founded Life Science Editors, where the team treats proposal editing as layered risk reduction across scientific, communication, and reviewer-cognition risk rather than proofreading. Life Science Editors offers a workshop discount to attendees, and Dr. Matson provided additional resources for the session.

Raphaël Bernier is Head of Growth at Atom Grants. He leads partnerships and hosts the webinar series connecting research teams with specialists across the funding landscape.