Accessible Webinar Guide: Planning to Recording

Why webinars need their own accessibility plan

A webinar combines several things that each have accessibility needs: a registration page, a live video platform, slides, spoken presentations, a chat and Q&A channel, and a recording. A gap in any one shuts some people out. A deaf attendee needs captions or an interpreter in real time, not a week later. A blind attendee needs to hear what is on each slide. Someone using only a keyboard needs to reach the Q&A box. Live delivery leaves little room to improvise, so the plan has to be made in advance.

Before: registration and access requests

Registration is the moment to find out what people need, while there is still time to arrange it.

Ask an open question such as: Do you need anything to take part fully? Give examples: live captions, a sign language interpreter, slides in advance, a phone dial-in option.
Say by when requests should be made so you can book services, and confirm each request personally.
Make the registration form itself accessible: labelled fields, keyboard operable, no image-only CAPTCHA without an alternative, clear error messages.
Tell people in the invitation what access is already planned, such as live captions for every session, so they do not have to ask.
Send the joining link and instructions in plain text, including how to turn on captions in the platform you are using.

Before: platform and slides

Choose the platform with accessibility in mind, and test it rather than trusting its feature page. Check that attendees can join and use chat, Q&A, polls and reactions by keyboard and screen reader; that captions can be displayed and resized; that an interpreter's video can be pinned or spotlighted so it stays visible during screen sharing; and that a phone dial-in exists for people with poor connections or who cannot use the app.

Slides need the same care as any document:

Large text, strong contrast and a simple layout, since slides are often viewed in a small window.
Few words per slide; detail belongs in speech and in the handout.
No information conveyed by colour alone.
Alternative text for images and charts in the shared file, and a correct reading order.
No flashing animations or rapid transitions.
Shared in advance in an accessible format, so screen-reader users and people who need extra reading time can follow along.

Before: live captioning and interpreters

Live captioning options differ in accuracy, cost and preparation. How the methods work technically, including latency and roll-up display, is covered in how live captioning works; the summary for planning is:

Human captioner (CART): A trained stenographer types captions in real time. Generally the most dependable for names, jargon and accents; needs booking ahead and preparation materials
Respeaking: A trained respeaker repeats the speech into recognition software tuned to their voice, adding punctuation as they go. Needs booking and preparation
Automatic live captions: Built into many platforms and available instantly. Quality varies with audio, accents, crosstalk and specialist terms, and errors are not corrected live

Whichever you choose, send the captioner or the platform's vocabulary settings the agenda, speaker names, slides and a glossary of product names and acronyms in advance. Automatic live captions can be a reasonable baseline for internal meetings, but for public events, or when an attendee has asked for captions, a human service is usually the safer choice.

Sign language interpretation is a separate service from captioning. Many Deaf people whose first language is a sign language follow an interpreter more easily than captions, and sign languages differ by country, so ask which one is needed. For long sessions, interpreters commonly work in pairs and take turns. Why speech-to-text cannot replace interpretation for Deaf signers is explained in sign language and AI translation.

Before: brief your speakers

Describe visuals aloud: say what a chart shows, not just look at this.
Say your name before speaking in a panel, at least for the first few turns.
Speak one at a time; crosstalk defeats captioners and interpreters alike.
Use a headset or a decent microphone in a quiet room. Clear audio improves live captions, the interpreter's work and the recording.
Avoid reading chat messages silently; read questions aloud before answering them.
Pace yourself, especially through technical terms and numbers.

During: running the session

Open with short housekeeping: how to turn on captions, where the interpreter is, how to ask questions by chat, Q&A or voice, and that the session is being recorded. Keep it brief and repeat it in the chat as text.

Keep the interpreter visible when someone shares their screen.
Read out every question you answer, and say the name of the person asking if they agreed to be named.
Describe poll questions and results aloud instead of only showing them.
Speak any link you share in chat, or say you will include it in the follow-up email.
Avoid relying on emoji reactions as the only way to respond.
Schedule a break in sessions longer than about an hour; captioners, interpreters and attendees all need it.

Hypothetical example: a 60-minute product webinar with three panelists

A software company asks at registration about access needs; two people request captions and one requests British Sign Language. The team books a human captioner and two interpreters, sends them the slides and a glossary of feature names a week ahead, and briefs the panelists to say their names before speaking. During the session the host reads every Q&A question aloud. Afterwards, the recording goes up with corrected captions, a transcript and the slides, and the follow-up email links to all three.

After: the accessible recording

Many registrants watch the recording instead of attending live, so it deserves the same care as the event itself.

Edit the recording: trim waiting time at the start and technical interruptions.
Produce captions for the edited recording. Live caption files rarely line up with an edited video and contain errors made under time pressure, so correct them or generate a fresh draft from the final edit.
Review the captions against the audio, fixing speaker names, product terms and numbers first, and add speaker labels where several people talk. The approach in how to proofread an AI transcript works for caption drafts too.
Prepare a transcript, including the Q&A, so people can read or search the session.
Publish the recording, captions, transcript and accessible slides together, and link them in the follow-up email.

Placing the transcript on the event page is covered in publishing transcripts alongside videos.

Trade-offs and common mistakes

Relying on automatic live captions for a public event and discovering that every product name came out wrong.
Booking an interpreter but letting the platform's screen-share layout hide them.
Speakers saying as you can see here throughout, leaving blind attendees with half the content.
Q&A that only works through a mouse-dependent widget.
Publishing the recording with unedited live captions, or with none.
Treating the recording as a substitute for live access; people who asked for captions needed them during the event.

Budget is a genuine trade-off. Human captioning and interpretation cost money and need lead time, which is why asking at registration matters: it tells you where to spend.

Where mydubly fits: the recording only

mydubly has no live mode, so it plays no part during the webinar itself. It fits the after stage. Export the recording from your webinar platform as a video or audio file, choose it from your device in mydubly's video to text tool, and you get a plain transcript, a timestamped transcript with [m:ss] labels, and SRT and VTT caption files. Files can be up to 2 hours long; a longer session needs splitting first. The tab must stay open until the job finishes.

The caption files contain recognized speech only, as single-line cues with no speaker labels and no sound tags, so a panel webinar needs names added and the whole draft reviewed before it counts as accessible captions. If the platform exports separate audio tracks per speaker, only one track is used, so upload the mixed recording. A 60-minute webinar costs 60 credits (6¢) to transcribe. The same files can be produced with the subtitle generator. Making versions in other languages is covered on the webinar translation page.

Next step: add an access question to your next registration form

Add one open access question and a request deadline to the registration form for your next webinar, and put the captioning arrangement in the invitation. Then plan the recording workflow before the session, so the edited captions and transcript are ready to publish shortly after the event.

Frequently asked questions

What should a webinar registration form ask about accessibility?

One open question works best, such as asking whether attendees need anything to take part fully, with examples like live captions, a sign language interpreter or slides in advance. Add a deadline for requests and a contact for questions. Then confirm each request personally so the attendee knows it has been arranged.

Are automatic live captions good enough for a webinar?

They are better than nothing and fine for many internal meetings, but accuracy drops with specialist terms, names, accents and crosstalk, and errors are not corrected during the session. For public events, or when an attendee has requested captions, a human captioner is usually the more dependable choice.

Do I need a sign language interpreter if I already have captions?

Possibly. Captions and interpretation serve different people: many Deaf attendees whose first language is a sign language prefer an interpreter, while hard-of-hearing attendees and others often prefer captions. Asking at registration tells you what your audience needs, and which national sign language is required.

How do I make webinar slides accessible?

Use large text, strong contrast and few words per slide, avoid conveying meaning by colour alone, add alternative text to images in the shared file, and avoid flashing animations. Share the slides in advance in an accessible format, and ask speakers to describe charts and images aloud during the presentation.

How soon should I publish the accessible recording?

As soon as captions have been reviewed, ideally within a few days so it reaches attendees while the topic is fresh. Publishing a recording without captions first and adding them later leaves out the people who most need them, so it is better to wait a little and publish everything together.

Can mydubly provide live captions during a webinar?

No. mydubly works only with recorded files chosen from your device, with no live mode or meeting integration. After the event, it can produce a transcript and SRT and VTT caption drafts from the recording, which still need review and speaker names before publishing as accessible captions.