Several times a year, my inbox fills with grant proposals from scientists who are seeking support for their work. I am the director of grants for a small family foundation that has supported research for more than half a century. My desk is the first step in a process that will determine the fate of these funding requests. I start by reading the proposal's title and the introductory summary of its significance.

Credit: Luciano Lozano/Getty

Then, I turn to the budget page.

The first thing I look at is the bottom line. If the full request is in our range, I skim through the rest with an eye for potential red flags. For example, we have a ceiling on overhead requests: we do not support operating funds or proposals that are primarily for equipment. Only after I have assessed whether there are issues with the budget do I return to the proposal itself.

When drafting a budget, it is crucial to know the maximum grant amount awarded by the foundation to which you are applying. If you ask for much more, the granter might reject your proposal on that basis alone. Some organizations avoid publishing an exact figure to discourage applicants from reflexively requesting it. How can you find out? In the United States, the Foundation Center, a national database for grant-seekers, provides access to the 'Form 990' tax return that all foundations are required to submit to the federal government (see go.nature.com/panexz). These returns disclose the dollar range of grants issued by a foundation each year. You should also look at the granters' websites. Most do not post their 990 forms, but some have annual reports or web pages that list project titles or recipients and the corresponding grants' dollar amounts. Be aware that when granters see budgets, for example, for exactly US$25,000 or $50,000, many will suspect that aspects of the budget were contrived to reach this exact amount. Proceed with caution.boxed-text

It is also important to find out whether a foundation has specific requirements for the budget section of grant applications — and, if so, to observe them strictly. Many do, and they typically post this information on their websites along with their application guidelines. These rules will tell you whether you can, for instance, purchase equipment, support undergraduate assistants or request publication fees for journal articles that may or may not arise from your work. They also specify whether there is a ceiling for overhead — or 'indirect' — expenses. With our limited resources as granters, we want to fund research, not depreciation and administration, so if an institution submits an application that proposes adding 50–70% — or even more — to the bottom line for overhead expenses, that can be a problem.

The budgeting rules will also tell you whether you can ask for general operating support or only specific project-related funds, and whether the funder will award multiple-year grants. Find out before you start budgeting. If your questions are not answered on the website, send an e-mail to the foundation to ask for clarification. There is nothing unusual about these enquiries.

The majority of the budget requests that we receive are indeed for our maximum grant amount, and many staff members from other grant-giving foundations have told me the same thing. Or, put another way, they have the same complaint. The place where we tend most often to see this escalation is in the budget's salary line.

Salary samples

We received an e-mail proposal recently that requested US$14,550, with more than half of that allocated to salary. When the hard copy of that proposal arrived, however, it looked as though the applicant had accessed our 990 form, because the salary allocation — in an otherwise unchanged proposal — had more than doubled. The applicant had added months' worth of support, raising the total request to $24,550.

We also received a proposal in which the principal investigator indicated that he had asked a second funder for $15,000, in addition to the $25,000 that he was seeking from us. I asked to see that second budget and discovered that he would actually require an extra $1,700 if he received only the smaller grant. That is, a total of only $16,700 would have allowed this preliminary investigation to proceed — in the absence of any salary support at all. He would have had to scale back his project, but its successful completion would have nonetheless enabled him to make a larger request for state or federal funding.

Although both of these researchers should have explained their position, I did not consider either case to be an act of duplicity (we funded one of them). An applicant's use of the salary line to increase a grant request is understandable to — if not welcomed by — the granter. There is prestige for a researcher in obtaining large grants, and some institutions do require staff to bring in salary support.

If a grant application has sparkle — novelty and a thoughtful protocol — it could persuade us.

You should visit your development office early in the budgeting process to talk to staff members and learn what they expect. Most universities and non-governmental organizations are prepared to reduce their overhead requirements to adhere to foundations' guidelines. We often see budgets in which the home institution produces matching funds to enable its researchers to stay within our support guidelines. If this is the case, consider adding that information to your budget page in a column labelled 'matching funds' or 'other sources' to show that your request is backed by a contribution to salary, fringe benefits, overhead expenses or any other support your home institution is willing to provide.

You might also want to include a column for other sources when you are applying to two or more granters for the same project. Here, it is important to explain in your budget justification whether the funds from these sources are already in hand (see 'Explain your budget'). If not, you need to clarify whether the project will still proceed — perhaps on a smaller scale — if the other organizations fail to contribute. State in your application when you will know about any other funds. The timing is important because a granter will be reluctant to write you a cheque if your project runs the risk of halting midway for lack of that support.

Too much information in a budget is better than too little. On my desk recently was a proposal from a principal investigator who wanted to travel to South America — with her dog. This added, yes, 13 lines to her budget, ranging from $30 for a flea-and-tick collar to $400 for a canine plane ticket. The researcher explained that her dog is not merely her travel companion but also a well-trained scat-detection dog — a necessity for her research. We funded this interspecies team.

If a grant application has sparkle — novelty and a thoughtful protocol — it could persuade us to work with the researcher to reduce the final request to better fit our resources. Or we may even agree to write a cheque that is larger than usual.