Short answer: Dictate grant proposals section by section, not front to back. Outline against the funder's review criteria first, then narrate each narrative section aloud in one pass without editing. Keep budgets, tables, and forms on the keyboard. Speech runs 130 to 150 words per minute, so a rough first draft arrives far sooner.
Grant writing has an awkward shape. The thinking is mostly done before you sit down: you know the program, you know the population you serve, you have argued for this work out loud in a dozen meetings. What is missing is the four thousand words that turn all of that into a submittable narrative. Most grant writers are not blocked on ideas. They are blocked on transcription of their own head, and that bottleneck runs at typing speed.
The average adult types around 40 words per minute. Practiced professionals reach 80 to 100. Ordinary speech runs 130 to 150 words per minute without any training at all. That gap is why dictation is worth learning for grant work specifically, where the deliverable is long, prose-heavy, and deadline-bound. This guide covers what actually works: which sections speak well, which do not, how to handle funder terminology, and how to keep a dictated draft rigorous rather than rambling.
Why grant narratives suit dictation better than most writing
Not all writing benefits equally from speech. Code does not. Spreadsheets do not. Grant narratives do, for three structural reasons.
The genre is explanatory. A needs statement, a program description, an evaluation plan, and a sustainability section are all forms of explaining something to a skeptical stranger. That is the same cognitive task as briefing a board member or answering a program officer's question on a call. You already have a spoken register for it.
The content is largely known. Unlike creative writing, a proposal is rarely discovered on the page. You know your service model, your staffing, your outcomes framework. Dictation is at its best when you are transferring known content, and at its worst when you are searching for content you do not yet have.
The volume is high and the deadline is fixed. A federal application can run twenty to forty pages across narrative, attachments, and forms. Foundation applications are shorter but often come in clusters, five or six at a time, each with a slightly different question set. Volume with a hard deadline is precisely where a 3x input speed difference compounds.
The catch is that speech produces a different kind of first draft than typing does. It is longer, looser, more repetitive, and more conversational. That is a feature if your revision process is strong and a problem if you were hoping to submit the first pass. Plan for a real editing round; you will still finish ahead.
What to dictate and what to keep on the keyboard
The single biggest mistake is trying to dictate the entire application. Sort the work first.
Dictates well
- Statement of need. Narrative, argumentative, and something you have said aloud before.
- Organizational background and capacity. Close to boilerplate, but each funder wants it angled differently. Speaking the angle is faster than editing a template into shape.
- Program description and activities. Naturally sequential. Walk through the participant journey out loud and you produce the section almost in order.
- Evaluation and data use narrative. The prose around the logic model, not the model itself.
- Sustainability, dissemination, and partnership sections. Short, formulaic, easy to talk through in a minute or two each.
- Cover letters and abstracts. Best dictated last, once you know what the proposal actually says.
- Budget narrative prose. The justification sentences, not the numbers.
Keep typing
- Budget tables and any spreadsheet work. Numbers, cell navigation, and formulas are keyboard tasks.
- Logic models and structured tables. Layout-driven, not prose-driven.
- Online form fields with strict character limits. Draft elsewhere, count, then paste.
- Citations and reference lists. Speaking a citation is slower than pasting one.
- Final compliance pass. Line-by-line checking against the funder's checklist is reading work, not writing work.
Roughly two-thirds of a typical narrative-heavy application falls in the first list. That is where the time goes and where the time is recoverable.
Setting up so dictation is available everywhere
Grant work is scattered across applications. The narrative might live in Word or Google Docs, the deadline tracker in a project tool, the portal in a browser tab that behaves like software from 2011, and the notes from your program director in whatever you had open at the time. A dictation tool that only works inside one editor will not survive that workflow.
Voice Keyboard Pro runs from the Mac menu bar and works system-wide. You hold a hotkey, speak, release, and the text appears at your cursor in whatever field is focused. That includes a grants portal text area, a comment in a shared doc, and an email to a program officer. There is nothing to open and no per-app configuration, which matters when a single afternoon touches eight different windows.
On the phone side, the iPhone keyboard has a built-in mic button, so the same dictation works in any iOS app. This is more useful than it sounds for grant writing. Site visits, conversations with program staff, and the sudden 9pm realization about how to frame the need statement all happen away from the desk. Capturing 200 spoken words into a note on the phone is the difference between having that paragraph and reconstructing it from memory on Monday.
Teach the tool your funder vocabulary first
Grant language is dense with acronyms, program names, and proper nouns that no general transcription system has strong priors for. Left alone, a dictated draft will contain a scattering of near-miss words that you then have to hunt down, and that hunting eats the time you saved.
Fix it once at the start. Voice Keyboard Pro's Smart Vocabulary is a personal dictionary with replacement rules, so you can add the terms you use constantly and have them come out correctly every time:
- Funder and agency names, including the ones that are initialisms in writing but spoken as words
- Your own program names, which are frequently invented compounds
- Notice and opportunity identifiers you reference repeatedly
- Partner organizations, school districts, county names, and neighborhood names
- Framework and instrument names used in your evaluation plan
- Staff names and titles that appear in the key personnel section
Twenty entries covers most of a typical portfolio. It takes about ten minutes and pays for itself inside the first proposal. If you want the longer version of how a personal dictionary changes accuracy on specialized work, we covered it in how custom vocabulary teaches dictation your words.
A section-by-section dictation workflow
This is the part that separates a usable draft from a transcript you have to rewrite from scratch.
Step 1: Build a compliance outline before you speak a word
Open the funder's guidance and turn the review criteria into headings, in the order the reviewers will read them. Include the point values if the funder publishes them. Under each heading, type a one-line reminder of what must be proved there. This outline is typed, not dictated, because it is structural work and it is short.
The outline is what keeps dictation honest. Speaking without one produces exactly the meandering draft that gives voice writing a bad reputation. Speaking with one produces a draft that maps to the scoring rubric, which is the only map that matters.
Step 2: Narrate one section per pass, without editing
Put your cursor under a heading, hold the hotkey, and explain that section as if a colleague had asked you to. Do not stop to fix words. Do not reread the previous sentence. If you lose the thread, release the key, take a breath, and start the next sentence fresh.
Two habits make this work. First, speak in complete sentences rather than fragments; the output is far cleaner. Second, say your punctuation where it matters for meaning, particularly periods at the end of long sentences and colons before lists. Most sections take two to five minutes of talking, which is a 300 to 700 word chunk of prose.
Step 3: Mark gaps out loud instead of stopping
You will hit a spot where you need a number you do not have memorized: the exact participant count, last year's completion rate, the current indirect rate. Do not go look it up. Say a marker, something like "T K figure here for completion rate," and keep going. Momentum is the entire value of the dictation pass, and one lookup can cost fifteen minutes to a browser tab.
When the pass is done, search the document for your marker and fill every gap in a single sitting with the source documents open. Batching lookups is faster than interleaving them, and it produces better sourcing because you are in verification mode rather than writing mode.
Step 4: Read it back and cut
Dictated prose is reliably 15 to 25 percent longer than it needs to be. Spoken sentences carry throat-clearing openers, hedges, and restated ideas. This is the predictable cost, and the edit is mechanical: delete the openers, collapse the repeated ideas, tighten the hedges. Reading the section aloud once more is the fastest way to hear what is redundant.
Small wording fixes do not require going back to the keyboard. Voice Edit lets you speak the change you want rather than clicking into the sentence, which keeps you in the same mode you drafted in.
Step 5: Do the compliance pass with fresh eyes
Last, go back to the funder's checklist and confirm every required element is present, page limits are respected, and formatting rules are met. Never do this immediately after drafting. The reason a dictated draft can fail review is almost never the prose quality; it is a missed requirement, and missed requirements are invisible to the person who just spent three hours talking.
Capturing the interviews that feed the proposal
Much of a strong proposal comes out of other people's heads. Program directors know the service model in operational detail. Frontline staff know the barriers participants actually face. Finance knows why the budget looks the way it does. That information usually arrives in a meeting and then decays.
Meeting Mode on the Mac captures those conversations with speaker detection and produces AI notes, which turns a 40 minute planning call into a set of attributed points you can pull from while drafting. For grant work the speaker attribution matters more than the raw transcript. Knowing that the outcome target came from the program director rather than from an executive's optimism changes how you write and defend it. Calendar meeting detection means the capture starts without you remembering to start it, which is the failure mode most note-taking tools never solve.
Use those notes as source material, not as draft text. Meeting language is fragmentary. The workflow is to read the notes, then dictate the section, not to paste the notes and edit them into prose.
Handling the parts that trip dictated drafts up
Numbers, percentages, and dollar figures
Speak numbers deliberately and confirm them in the edit pass. Grant narratives are full of figures that must match the budget exactly, and a mismatch between narrative and budget is one of the more common technical review findings. Treat every dictated number as unverified until you have checked it against the source. This is not a dictation weakness so much as a discipline that grant writing demands regardless of input method.
Quotations and cited claims
Do not dictate a quotation from memory. Mark the spot, paste the exact text later. The same goes for any claim that will carry a citation. Speaking generates fluent sentences quickly, and fluent sentences are exactly the ones that slip past your own review with an unsourced statistic embedded in them.
Character-limited fields
Many online portals cap sections at a character count rather than a word count. Dictate into a document where you can see a live count, trim to fit, then paste. Dictating directly into a portal field that silently truncates is a bad afternoon.
Sounding conversational when the funder wants formal
Spoken register is more casual than proposal register. The fix is not to speak formally, which slows you down and sounds stilted anyway. The fix is a targeted edit pass for register: replace contractions, remove second person where the funder prefers third, and convert the two or three places where you got chatty. This takes a few minutes per section and preserves the speed advantage of natural narration.
What this does to your timeline
Be realistic about the arithmetic. Grant writing is not purely an input-speed problem. Reading the guidance, gathering data, coordinating letters of support, and assembling attachments all take the same time regardless of how you write. Dictation compresses one specific stage: getting the narrative from your head into a document.
For a narrative that runs 4,000 words, typing at 40 words per minute with normal pausing to think is a multi-session task. Narrating the same content in five-minute section bursts is closer to an afternoon, plus a real editing round the next morning. The draft you get is rougher. The draft you get also exists, which is the state most late applications never reach.
The second-order effect matters more for most teams. When a first draft is cheap to produce, you stop protecting bad drafts. You are willing to throw out a needs statement that is not landing and talk through a new one, because the cost of the new version is ten minutes rather than an evening. Grant quality improves more from a discarded draft than from a defended one.
Privacy considerations for sensitive proposals
Grant narratives regularly contain information you should be careful with: participant demographics, unpublished program data, partner arrangements that are not public yet, and salary detail in the budget narrative. It is a fair question to ask what happens to the words you dictate.
Voice Keyboard Pro's servers store only operational pings. No audio and no transcript content is retained server-side. Your dictated narrative goes to your cursor and lives in your document, and your usage history stays on your machine. If your organization has a policy about where draft program data may be processed, that is the detail to bring to the conversation.
Frequently asked questions
Will reviewers be able to tell a proposal was dictated?
Not after a proper edit pass. They can tell if you skip it. Unedited dictation reads as loose and repetitive, which reviewers register as weak writing even if they do not identify the cause. Edited dictation reads like clear, direct prose, which usually scores better than the over-constructed sentences people produce when typing slowly and revising as they go.
Is this useful for short foundation applications?
Yes, arguably more so. A 500-word letter of inquiry is a single spoken pass of about four minutes. The efficiency gain per application is smaller, but the number of applications is larger, and short applications are the ones organizations skip because there is no time. Making them cheap changes which opportunities you can pursue.
Can I dictate the whole thing on my phone?
You can, and it works for capture and for short sections. For a full narrative, the Mac is the better environment because you need the outline visible, the guidance open, and a real editing surface. The practical split is phone for capture and fragments, Mac for drafting and revision.
Does it handle academic and research proposals?
The same workflow applies to research grant narratives, with heavier reliance on the personal dictionary for methods terminology and instrument names. If you are writing in an academic context, the related workflows in voice typing for PhD students and voice to text for researchers cover the literature-heavy side in more depth.
Getting started on your next application
Do not convert your whole process at once. Pick one section of the proposal currently on your desk, ideally the organizational background, since you know it cold and the stakes are low. Build the outline, add ten terms to your dictionary, and narrate that section in a single pass. Time it. Then edit it and time that too.
Most people find the drafting stage collapses and the editing stage grows slightly, with a large net gain. If that holds for you, extend it to the needs statement and program description on the next application, which is where the bulk of the writing lives.
The proposals that do not get submitted are almost never the ones that lacked a good idea. They are the ones where the narrative never got written down in time.
Voice Keyboard Pro has a free tier with daily limits, which is enough to test the workflow on a section or two. Pro is $4.99 a month or $34.99 a year and removes the limits, which is what you want during a submission week when the narrative is long and the deadline is not moving.