Your proposal is not read in a quiet room by someone with an open afternoon. It is read by a working scientist or engineer who agreed to review a stack of them, usually in the evening, alongside a day job, and yours may be the eighth in the pile. Writing for those conditions earns points that have nothing to do with how good the technology is.

Assume two passes, and design for the first

Almost every reviewer reads twice: a fast pass to work out what you are proposing, and a slower pass to fill in a score sheet. Proposals get lost on the fast pass. A few habits survive it.

  • Headings that state content, not category. Aim 2: Demonstrate 500-hour electrode stability beats Technical Approach, Part 2. A reader who only reads your headings should still be able to describe your project.
  • Topic sentences that carry the point. The first sentence of a paragraph should be the one you would keep if the rest were deleted. Put the conclusion there and the supporting detail after it, never the reverse.
  • Short paragraphs. Three to five sentences. A half-page block gets skimmed rather than read, and whatever was buried inside it does not get scored.
  • One idea per paragraph. Two ideas fused together means one of them is invisible.

Remove the friction in your language

Define every acronym at first use, and define it again at the start of each major section. A reviewer who reads your significance section on Monday and your approach on Thursday should not have to page backward to remember what your three-letter subsystem is. Cut jargon that is local to your subfield: panels are assembled to cover a topic area, not your exact specialty, and the person scoring your innovation may be two steps away from your literature.

Attach units to every number and state what a number is compared against. Adjectives get discounted automatically; novel, revolutionary, and game-changing are treated as noise, while a measured value against a named benchmark is treated as evidence. If you find yourself writing that something is highly innovative, replace the sentence with the thing that makes it so.

Put each claim where the scorer looks for it

The evaluation criteria are published in the solicitation, and they usually map to something like significance, technical approach and feasibility, team and facilities, and commercial potential. Answer each criterion under a heading that visibly corresponds to it. Burying your commercial argument inside a technical section is a reliable way to lose points you had already earned, which is why the commercialization plan gets its own treatment rather than being sprinkled through the text.

Before you submit, read the criteria list and your own document side by side, and make sure each criterion has an obvious home. Reviewers who cannot find an answer generally record that it was missing, not that it was elsewhere. It helps to understand how a review panel actually scores before you finalize the structure, and to reread the general SBIR application requirements for the formatting rules that get proposals returned unread. If writing under those constraints is not your strength, that is also part of the calculation when you decide whether to hire a grant writer.

Projects House writes the technical content behind SBIR proposals for hardware and device companies, in language a mixed panel can score. Send us your draft or your idea through the contact form.