• See the Presentation on How to Become a Successful NSF PI (PowerPoint)

  • Disclaimer -Any opinions, findings, and conclusions or recommendations expressed in these slides are those of the authors/presenters and do not necessarily reflect the views of the National Science Foundation.

Whether you have a firm idea of the research direction you want to pursue or not, following the steps below will help you home in on a competitive research direction.

• Read program solicitations to find the ones that seem closest to your research interests and goals.
• Think about how your research interests and goals could support the vision and goals of the solicitations. (You may have to morph your research goals a bit or take some liberties with stated program goals.)
• For the solicitations that fit you best, read all the abstracts of projects funded over the last two years and see if you can find an important gap that your research could fill.
• Now it is time to talk to the cognizant program directors of the programs that fit you best. If your best fit is Cyber Human Systems, talk to me or other program directors in that cluster. For other programs, you will have found the relevant program directors during the previous steps. Contact them directly. Not only will this discussion help you fine-tune your plan, but it's an important way to help research programs evolve to maintain their relevance and vitality, and thus is an important contribution to the research community.
• In cases where the intellectual contributions of your project are significant with respect to two or more programs, you should contact program officers in both programs to discuss your project goals. If the associated program managers see the value of your project to their programs and know of the other program mangers' interests, they could agree to jointly fund a project that none of their programs would or could fund alone.

• Answer these questions clearly and concisely in the project summary and introduction: (1) What is the problem? (2) Why is it important to solve? (3) Which parts of the problem will you solve? (4) How will those problems be attacked? The second question points to potential broader impacts of the research. The third and fourth point to the core intellectual contributions, which help a program manager find the most appropriate reviewers.
• Try to appeal to wide audience, since the breadth of topics and technical approaches appearing in a single panel can be very wide. An application with a clear positive societal impact (even if 10 years out), an insightful discussion of the most relevant previous work and on-going research projects, and a prediction of the technical impact the research could have on the field are extremely important.
• To emphasize the importance of questions (3) and (4) above, clearly state in the proposal summary and introduction, the area in which your primary intellectual contribution will be. Some proposers write Intellectual Merit statements that are impressively broad, but obscure the area of the primary contribution, which makes it hard for program directors to identify the most appropriate reviewers.
• Demonstrate total command of the relevant literature by citing the earliest key results, not only the most recent relevant results. Also you should avoid writing, "...to the best of my knowledge..." This just draws attention to the fact that you believe you might not be on top of all the relevant literature. If you think you can't avoid it, go back to the literature.
• Never dispense with a large swath of the literature by simply saying, "Most research on topic A takes approach X, but I will use approach Y." What's important here is to make clear that your approach is worthy of pursuit in comparison to all other existing approaches, not just the majority (possibly, misguided) approach. If you're going to use a statement like this to give perspective, follow it up with statements that cover the minority approaches too, since they could be the current best and most similar to yours.
• Do not simply pose research challenges in your research plan. Also offer plausible ways to tackle them.
• If your proposal is not funded, the reviews can be very valuable in formulating a resubmission. There is often both a "consensus" panel review, and individual reviews. The individual reviews may include comments that others did not agree with, and which are not reflected in the consensus review, but the individual reviews may also have more specifics for you to consider. Your program director may be able to help you interpret the reviews.
• One of the most valuable experiences for writing proposals is serving on a review panel with other reviewers. Tell your program director that you are interested in doing this, either in his/her program or another one.
• Ask your senior colleagues to let you read their successful proposals, and maybe even their reviews. Also ask them to read your declined proposal and reviews to help you understand how to make it competitive.

Some advice from others

• Twelve steps to successful proposal writing by long-time NSF program manager, George Hazelrigg.
• More from George H.

• To make the best use of your time and that of the program director, do your homework, that is, carry out as many of the steps for identifying a fundable research project as possible.
• Prepare a 1-page white paper w/short answers to the following questions before talking with a program director: 

(1) What is the problem?

(2) Why is it important to solve?

(3) Which parts of the problem will you solve and how?

(4) What is the intellectual merit of your proposal?

(5) How does this fit Cyber Human Systems (CHS) program?

1. If you don’t hear back on your first contact (phone or email), what is the best next step and where is the line between persistence and annoyance?

It is generally suggested that you start with an e-mail, because many PDs are out of the office for various reasons (meetings, panels, conferences, independent research activities, etc.). PDs are very busy so give them a week to respond, many will respond within a day or two. If the PD does not respond after 2x, then try to contact another PD within that program. It is recommended that you contact one PD in the program, instead of all PDs in that program. By sending a note to all it often times because unclear who should respond, risking the chance of it falling deeper in their mailbox, resulting in no response at all.

2. How much about your grant idea should you share when reaching out via email? 1 page, 2 pages? What should that info include?

Because time is of essence most PDs will only want a one-page summary, definitely no more than two pages. They always work to treat everyone fairly, so what they do for they must do for everyone. This is why they cannot read an entire 15-page proposal to provide feedback, there is not enough time in doing so for everyone. In your project summary, you should strive to answer these questions clearly and concisely: (1) What is the problem? (2) Why is it important to solve? (3) Which parts of the problem will you solve? (4) How will those problems be attacked? (5) How does this fit the particular program of interest? Pertaining to question #5, this tells the PD whether you have done the homework necessary to determine the proper fit for the idea. Or it may help them in providing “fit” advice. That is many times a proposer may think their idea fits a certain program, division, or directorate but often times it doesn’t and this is where the PD can help. Work with mentors who have been successful on proposals to NSF, seek out those who will give you honest and constructive feedback, rather than those who simply say “looks great, submit”. Critical feedback from more experienced folks is critical to success at NSF!

3. What are some of the best questions you have had/asked? Don’t be shy to bring forth “crazy” ideas; NSF is the place for such ideas!

NSF is one of the only agencies that funds work on potentially “crazy” ideas that may lead to transformative ideas in terms of intellectual merit and broader impacts.

4. Are there any questions you would recommend NOT asking?

Gather what your interests and passions are and write for that. Don’t ask the PD what they would like funded and what their interests are in and write to that. PDs, for the most part, were once/are academics so they understand the granting process, funding students, publishing students, tenure process, etc. They have a very good understanding of the university system and how to navigate that path. PDs encourage communication, there will never be any negative consequences to speaking with PD, e.g. won’t hurt your chances for future funding etc. Feel comfortable asking them questions. Seek them out early in the grant writing process to ensure your idea is a good fit and if not where a good fit may be.

5. What is the goal of the program officer in these conversations? What is the goal of the faculty member?

The goal of the faculty member is to of course secure funding to fund your research, students, obtain tenure so forth and so on. The NSF PD knows this but what they want to hear is what it is that you are excited and passionate about in a clear and concise manner. PDs may not have expertise in your area, most likely they will not, so you will need to be generalists to some degree when conveying your idea. The PDs role is to encourage you to submit assuming it’s a good fit for their program or to help you navigate where may be the better home. As a proposer realize that you will be declined, the funding rate within my own program is from ~6% (CHS: LARGE) ~25% (CHS:CRII) depending on the solicitation. As a new faculty member write your PD and ask them if you can serve on a panel (send an e-mail with your CV attached) so that you can understand the review process. It is OK if you have never served or have not been successful via NSF funding mechanisms.

(Thank you to PDs in Cyber-Human Systems Program!)