How to Dictate in DocuSign on Mac and iPhone

Short answer: DocuSign has no dictation button of its own, so you dictate with a system-level tool. On Mac, click into any DocuSign field, hold your dictation hotkey, speak, and release. On iPhone, tap the field and use a keyboard with a built-in mic button inside the DocuSign app.

Most people think of DocuSign as a place where nothing gets typed. You upload a document, drop signature boxes on it, and send. The typing happened somewhere else, in the word processor, before the file ever arrived.

Anyone who actually sends agreements for a living knows that is wrong. A single envelope can absorb a surprising amount of writing: the subject line and the message to signers, the recipient names, the text fields you fill in on their behalf, the private note attached to one signer, the reason you are voiding an earlier version, the comment explaining why clause seven changed. None of it is long. All of it is fiddly. And it arrives in bursts, usually at the end of a deal, when you are tired and in a hurry and least inclined to type carefully.

That combination, short bursts of careful text under time pressure, is where dictation earns its keep. This guide covers how to do it properly on both Mac and iPhone, which fields are worth dictating, which ones you should never dictate, and how to keep the accuracy standard that contract text demands.

Why DocuSign has no dictation feature

It is worth understanding the gap before you try to fill it. DocuSign is delivered to you either as a web application in your browser or as a mobile app. Neither of those is the right place for a voice feature to live.

In the browser, any in-page dictation button would depend on browser speech APIs that behave differently in Chrome, Safari, Firefox, and privacy browsers, and would need a microphone permission granted per site. In the mobile app, text input belongs to the keyboard, which is deliberately outside the app's control. So DocuSign, sensibly, does not build one. Voice input is expected to come from the layer below: your operating system, or a tool that plugs into it.

This is the same structural reason dictation is provided by the system rather than the app in most web-based business software. If you have ever looked for a mic button inside a web form and found nothing, the pattern will be familiar, and it is covered in our guide to dictating into Google Forms.

The practical consequence is good news. Because voice input comes from below, one tool covers DocuSign, your email client, your CRM, and the document you drafted the agreement in, with no per-app setup.

Dictating in DocuSign on a Mac

The workflow is the same regardless of whether you use DocuSign in Chrome, Safari, or another browser, because you are not asking the browser for anything.

1. Open the envelope you are preparing or reviewing.
2. Click into the field you want to fill so the text cursor is blinking in it.
3. Hold your dictation hotkey.
4. Speak the text, including punctuation where you want it.
5. Release the key. The text appears at the cursor.

With Voice Keyboard Pro, that hold-speak-release gesture works in every field DocuSign presents, because the app types at your cursor system-wide rather than negotiating with the page. It lives in your Mac's menu bar, so there is no window to bring forward and nothing to copy across. The same hotkey works when you switch to the email confirming the send.

Two habits make this much smoother in practice. First, tab between fields rather than reaching for the mouse, so your hand stays on the keyboard and the hotkey stays under your thumb. Second, dictate the whole value in one pass rather than in fragments. Dictation systems use surrounding words to resolve ambiguity, so "net thirty from the invoice date" transcribes more reliably as one phrase than as three hesitant pieces.

Which fields are worth dictating

Not every box in an envelope benefits equally. The ones where voice genuinely saves time share a trait: they hold sentences rather than tokens.

• The message to signers. This is the highest-value field in DocuSign for dictation. It is the one place you write a real paragraph, it sets the tone of the request, and it is the field most often left as boilerplate because typing something better feels like too much effort. Speaking three specific sentences takes about fifteen seconds.
• Private notes to an individual recipient. Short, contextual, often explaining an exception. Perfect dictation length.
• Comments during negotiation. Explaining a redline in your own words is exactly the kind of writing that comes out better spoken than typed, because you would explain it out loud to the person's face anyway.
• Reasons and internal annotations. Voiding an envelope, declining, correcting, or logging why a term changed. These are the notes your future self needs and your present self writes badly because they feel like overhead.
• Long text fields you complete on a signer's behalf. Descriptions of scope, property details, deliverables, addresses.
• Template descriptions. Written once, read for years by everyone on your team.

Which fields to leave alone

Be equally disciplined about the other direction. Some fields should be typed, or better still pasted from an authoritative source.

• Signature and initial fields. These are not text fields. They are adopted signature blocks, and voice has no role there.
• Email addresses. A dictated address will get the domain or the punctuation wrong often enough that checking costs more than typing. Copy them from the contact record.
• Legal entity names on a first use. Get the exact registered name from the source document, including whether it ends in Inc, Incorporated, LLC, or Limited. Once it is correct, save it to your vocabulary so future dictations produce it correctly.
• Account numbers, tax identifiers, and reference codes. Long digit strings with no linguistic context are the worst case for any transcription system and the worst case for a human proofreader too.

The rule underneath all of this: dictate meaning, type identifiers.

Dictating in DocuSign on iPhone

The mobile app is where a lot of envelope work actually happens, because approvals arrive when you are between things. It is also where typing is most painful, which makes it the better argument for voice, not the worse one.

On iOS, dictation comes from the keyboard, so the method is:

1. Open the envelope in the DocuSign app.
2. Tap the field you want to fill. The keyboard appears.
3. Switch to a keyboard with a built-in mic button if it is not already active.
4. Tap the mic, speak, and tap again to stop.
5. Review the text in the field before moving on.

Voice Keyboard Pro's iPhone app is a custom keyboard with the mic button built in, so it works inside DocuSign the same way it works in Mail or Slack. Because it is a keyboard rather than an app feature, there is nothing for DocuSign to support and nothing to configure per app.

Three iPhone-specific features matter more here than they do elsewhere:

• Voice Edit lets you fix text by speaking the change instead of trying to place a cursor precisely in a small field on a phone. If a message to signers came out with the wrong date, you say what should change rather than thumbing through the sentence character by character. On a contract field, that is the difference between fixing it now and promising yourself you will fix it at the desk.
• Two-way translation while dictating covers 24 languages, which matters when the counterparty reads a different language than you speak. You can dictate a signer message in your own language and send it in theirs.
• Swipe typing is there for the fields and the moments where speaking out loud is not an option, which for contract work is often: on a train, in an open office, in a waiting room.

That last point deserves emphasis. Agreement text is frequently confidential. Dictation is the fastest input method available to you and also the least discreet one. Knowing when not to use it is part of using it well.

Getting names, numbers, and dates right

Contract text has a lower tolerance for error than almost anything else you write. An awkward sentence in a Slack message costs nothing. A wrong figure in an envelope that gets countersigned costs a renegotiation.

So treat the following as non-negotiable habits rather than tips.

Read back every figure before you send

Not the whole message. Just the numbers, the dates, and the names. Say them aloud against the source document. This takes about ten seconds per envelope and catches the entire class of error that matters. Speaking numbers is generally more reliable than thumb-typing them on a phone, but the consequence of a mistake is high enough that verification is worth the ten seconds either way.

Dictate amounts in the form you want them

Currency and date formats are where dictated text most often needs cleanup. Say the value the way it should appear, and check the result the first few times to learn how your tool renders it. Once you know the pattern, it becomes automatic. We go deeper on this in our guide to dictating numbers and dates accurately, which is worth reading once if figures are a daily part of your work.

Never dictate a name you have not verified

Personal and company names are the highest-risk words in any agreement, because they are exactly the words a general transcription engine has never seen. Verify once from the authoritative source, then make the tool remember it.

Build a vocabulary for the language you actually use

This is the step that turns dictation from usable to genuinely fast for agreement work, and it is the step most people skip.

Every deal has a small, stable set of terms that recur in every envelope: the counterparties you work with repeatedly, your own entity name, your product names, the clause vocabulary of your industry, the abbreviations your team uses internally. A general-purpose transcription engine will guess at these, and it will guess consistently wrong in the same handful of places, which is maddening precisely because it is predictable.

Voice Keyboard Pro's Smart Vocabulary is a personal dictionary with replacement rules built for this. You add the terms and the corrections once, and the output stops needing the same fix every time. For contract work, a useful starting list looks like:

• Every counterparty entity name in its exact registered form.
• Your own entity and product names.
• The names of the people who sign on your side and theirs.
• Standard clause shorthand your team says out loud.
• Any term your transcription has already gotten wrong twice.

That last rule is the practical one. Do not try to build the list up front. Add an entry the second time a word comes out wrong, and within two weeks the vocabulary reflects your real work rather than your guess about it. Our post on how a custom vocabulary learns the words you actually use covers the mechanics.

If you work in a regulated field with heavy terminology, the same discipline applies more strongly. Our guide to voice to text for lawyers goes through vocabulary building for legal language specifically, including how to handle terms that must be reproduced exactly.

A realistic workflow, start to finish

Here is what an envelope looks like when voice is doing the work it is good at and staying out of the work it is not.

1. Prepare the document elsewhere. Draft the agreement in your word processor, dictating the prose sections and typing the defined terms and figures.
2. Upload and add recipients. Paste email addresses from your contact records. Type or paste names on first use.
3. Dictate the subject and message. Hold the hotkey and say what the signer actually needs to know: what this is, what you need from them, and by when. Three sentences.
4. Place fields and fill any you complete yourself. Dictate descriptive text. Type identifiers.
5. Dictate private notes to individual recipients where context helps.
6. Read back names, figures, and dates against the source.
7. Send, then dictate your internal log entry while the details are still in your head. This is the step everyone intends to do and almost nobody does, purely because typing it feels like a chore after the deal is done.

The time saved is not dramatic on any single field. It is dramatic in aggregate, across every envelope, mostly because the fields you would have left blank or left as boilerplate now get filled properly.

The speed argument, briefly

The numbers behind all of this are unremarkable and hold up well. Comfortable speech runs at roughly 130 to 150 words per minute. The average adult types around 40, a professional typist around 80 to 100, and thumb-typing on a phone is usually slower than any of those. For a hundred-word signer message, that is a difference measured in minutes per day once you count every envelope.

But speed is not really the reason to dictate in DocuSign. The reason is that the fields which improve an agreement's chances, the clear message, the note that pre-empts a question, the comment that explains a change, are all optional fields. Optional fields get skipped when they are expensive to fill. Lowering their cost is what changes the behaviour.

Dictation does not just make you faster at the writing you already do. It makes you willing to do the writing you were skipping.

A note on confidentiality

Agreement text is often the most sensitive writing you do all week, so it is fair to ask what happens to it. Before you adopt any voice tool for contract work, get a clear answer to a simple question: what leaves your machine, and what is stored.

For Voice Keyboard Pro, the answer is that our servers store only operational pings. No audio and no transcript content is kept on our servers, so the text of the agreement you dictated is not sitting in anyone's database. If you are evaluating alternatives, ask them the same question and expect an answer that specific.

Frequently asked questions

Does DocuSign have a built-in dictation or voice typing feature?

No. Text input in the DocuSign web application and the mobile app comes from your operating system or your keyboard, so dictation is supplied by a system-level tool rather than by DocuSign itself.

Can I dictate directly into a signature field?

No, and you should not want to. Signature and initial fields are adopted signature blocks rather than text inputs. Voice input applies to text fields, messages, notes, and comments.

Does it work in the DocuSign web app in any browser?

With a system-level dictation tool, yes, because the text is typed at your cursor rather than requested from the page. That also means it keeps working in privacy browsers where in-page voice features often fail. If you use one of those, our post on dictation not working in Brave explains why browser-level voice input breaks there.

What about the DocuSign iPhone app specifically?

It works the same as any other iOS app, because a custom keyboard is available wherever the system keyboard is. Tap into the field, use the mic button on the keyboard, speak, review.

Is dictated text accurate enough for contracts?

For prose, yes. Modern AI transcription handles normal sentences reliably. For identifiers, figures, and legal entity names, the right standard is not "accurate enough" but "verified," and that is true whether the text was spoken, typed, or pasted. Read back the numbers every time.

Can I use one tool for DocuSign and everything else?

That is the point of a system-level approach. The same hotkey on Mac and the same keyboard on iPhone cover DocuSign, your email, your CRM, and the word processor where the agreement started, with one vocabulary shared across all of them.

Getting started

If you send agreements regularly, start with the single highest-leverage change: dictate the message to signers on your next five envelopes instead of leaving the default. It is the field that most affects how quickly people sign, it is the field most often neglected, and it takes fifteen seconds to speak properly.

Then add the vocabulary entries as errors show up, and keep the read-back habit for numbers. Voice Keyboard Pro has a free tier on both Mac and iPhone, so you can run that experiment on real envelopes before deciding whether voice belongs in your agreement workflow permanently.