In short
- Clients often open your link where they can't play sound: in a meeting, on a train, next to a sleeping child. Subtitles let them watch anyway.
- Captions for prerecorded video are a Level A requirement in WCAG, the accessibility guidelines many organizations follow.
- Subtitles help clients who work in a second language follow names, numbers and technical terms.
- Speech-to-text gets most words right and some names wrong: speak clearly, show tricky names on screen and check before you send.
- On-device transcription creates subtitles without sending your voice to a third-party service.
Subtitles for videos you send clients mean they can watch even when they can’t turn the sound on: in a meeting, on a train, in an open office. They also help clients who work in a second language and make your video accessible to people who are deaf or hard of hearing. With on-device speech-to-text they cost you almost nothing to add.
Think about where your client opens your link. Probably not at a quiet desk with headphones on. More likely on their phone, between two other things, with people around. A video they can’t hear is a video they’ll “watch later”, which often means never.
Where clients actually watch
Client videos are short and arrive in the middle of the day, in an email or a chat. The client opens the link where they are:
- In a meeting, glancing at their phone while someone else talks.
- On a train or a bus, without headphones.
- In an open office where playing audio out loud feels rude.
- At home in the evening, with someone asleep in the next room.
In each case the choice is: watch on mute, or postpone. With subtitles, watching on mute works, because in a screen recording the screen shows what you’re doing and the subtitles carry what you’re saying.
Subtitles are an accessibility basic
The Web Content Accessibility Guidelines (WCAG) are the W3C’s standard for accessible web content. Success criterion 1.2.2, Captions (Prerecorded), is at Level A, the most basic level. It asks that captions be provided for all prerecorded audio content in synchronized media, so that people who are deaf or hard of hearing can follow it.
A private video to one client isn’t a public web page, and WCAG doesn’t govern your inbox. But the reasoning holds. If someone on your client’s team is deaf or hard of hearing, a video without subtitles excludes them from the conversation. Adding them is the polite default.
Clients who work in a second language
Many agencies work across borders. A client who reads English well may still find fast spoken English harder, especially with accents, technical terms and numbers.
Subtitles let them read along, catch the word they missed, and pause on a sentence. Prices, dates and names become much harder to mishear. If you speak to your client in their language, subtitles in that language help the same way.
Get the words right
Automatic subtitles are good at everyday speech and less good at the words that matter most in client work: names, brands, product terms and abbreviations. “Webflow” can become “web flow”, a client’s surname can become a common word, “CMS” can come out as anything.
To reduce mistakes:
- Record clear audio. Speech-to-text is only as good as what it hears. A quiet room and a close microphone help more than anything else; see mic setup for client videos.
- Speak at a steady pace. Fast, mumbled sentences produce fast, garbled subtitles.
- Show tricky names on screen. If the name is visible on the page while you say it, a subtitle error is easy for the client to read past.
- Say abbreviations in full the first time: “the content management system, the CMS”.
- Watch once with the sound off before you send. If the subtitles alone make sense, the video works on mute.
If a key name or number comes out wrong and you can’t fix it, mention it in the message you send with the link.
Before you send a subtitled video
- Names of people, brands and products are recognizable.
- Prices, dates and quantities match what you said.
- Abbreviations are spelled out once.
- The video makes sense with the sound off.
- The first sentence says what the video is about.
Where your voice goes when you make subtitles
Many tools create subtitles by sending your audio to a cloud speech-to-text service. For a product demo that’s fine. For a video discussing a client’s unreleased site, pricing or internal problems, you may prefer the audio stay on your machine.
On-device transcription runs the speech-to-text model on your Mac. The audio stays on the computer, and nothing about the client project reaches a transcription provider.
| Cloud speech-to-text | On-device speech-to-text | |
|---|---|---|
| Audio stays on your Mac | ✗ sent to a provider | ✓ |
| A third party hears the client project | ✓ the provider | ✗ |
| Fits | Product demos, public content | Unreleased sites, pricing, internal problems |
We compare both approaches in on-device vs cloud transcription, and if GDPR matters to your clients, our page on a GDPR-friendly screen recorder covers the rest.
For step-by-step options on a Mac, including the built-in tools and subtitle files, read how to add subtitles to a screen recording on Mac.
Subtitles are one part of getting watched
Subtitles remove one reason to postpone a video. Others are the message you send with the link, a clear title, and a first sentence that says why it matters. Our guide on getting clients to actually watch your videos covers those, and the guide to async video for client communication puts them in the bigger picture.
Sources
- W3C: Understanding Success Criterion 1.2.2, Captions (Prerecorded), WCAG 2.2, checked November 2026.
Frequently asked questions
Do client videos need subtitles?
They're not legally required for most private client messages, but they make a real difference: clients can watch without sound, follow more easily in a second language, and people who are deaf or hard of hearing can understand the video.
What's the difference between subtitles and captions?
Subtitles usually transcribe or translate speech. Captions also describe other meaningful sounds, like music or an alert tone, for viewers who can't hear the audio. In a talking-over-screen client video, the speech is nearly everything, so the two are close.
How accurate are automatic subtitles?
With clear audio, modern speech-to-text gets most words right. It struggles most with names, brand names, abbreviations and jargon, and with noisy recordings. Check those words before you send.
Can I create subtitles without uploading my voice?
Yes. Some apps run the speech-to-text model on your Mac, so the audio never leaves your computer. That matters if your videos discuss client projects you'd rather not send to a third-party AI service.



