Two things follow that change how an application is written
The first is that the review criteria are published in the same document as the instructions. An applicant is told, in advance and in writing, what the application will be scored against.
The second is that eligibility and preparation instructions are separate required elements, and meeting one does not imply meeting the other.1 An eligible applicant can be set aside on preparation alone.
Both facts point the same way. The notice is not background reading; it is the specification, and the application is a response to it.
The trace is the work
The single most useful document in a grant application is not part of the application, and no funder ever sees it.
Every requirement in the notice becomes a row, and every row gets the page and paragraph of the application that answers it. Rows with no answer are the finding. Rows answered twice, in two places that say different things, are the second finding.
That table turns a subjective argument about whether the application is strong into an arithmetic one about whether it is complete. It also survives a change of writer, which matters on an application assembled by several people under a deadline.
Editors Canada's copy editing standards cover the surrounding discipline. They name checking for missing information, flagging places that need citations, and resolving variations in terminology, logic, and mechanics.2
Where applications lose points without losing the argument
Four faults recur, and each is visible without knowing the program. All four cost points from a reviewer who agrees with the proposal.
Language that does not match the notice is the first, and a style sheet is where the matching gets recorded. An applicant who calls something a service and the notice calls it an activity has made a reviewer translate, and a reviewer with forty applications translates once and then stops.
Unstated denominators are the second. A number that describes a proportion of something the application never names cannot be evaluated, only believed.
Uncosted claims are the third. A promised output with no corresponding line in the budget invites the question of how it will be produced.
Inconsistency between the narrative and the attachments is the fourth, and it is the most common. A budget narrative, a logic model, and a work plan written by three people will disagree about scope unless somebody reads them together.
Clarity is a scoring property here
Reviewers read many applications in a short time, and the ones they can follow are the ones they can score. The federal definition of plain writing describes the target: writing that is clear, concise, well-organized, and follows other best practices appropriate to the subject and intended audience.3
Editors Canada's stylistic standards translate that into work an editor can do. They ask for wording made precise, ambiguity resolved, and wordiness eliminated by deleting redundancies and empty phrases. They also ask for plain language principles, with concrete terms suited to the reader.4
None of that changes what an organization proposes. It changes whether a reviewer can tell what is proposed on one reading, which is the only reading most applications get.
Where an editor stops
An editor does not decide whether an approach is sound, whether a budget is reasonable, whether an organization is eligible, or whether a claim about outcomes is supportable. Those are the applicant's determinations, and they arrive with the copy.
Everything a client sends is treated in confidence, including draft applications, budgets, and correspondence with a funder. We work on grant and nonprofit material and on project proposals against whatever the notice requires.
The most useful thing an applicant can hand an editor is the notice rather than the draft. A draft can be read against a notice. A notice cannot be reconstructed from a draft, and the same is true of a government solicitation.