Captions and Transcripts for Your Online Course
You spend a weekend filming a lesson you are proud of. You upload it, hit publish, and move on. But three people cannot fully use what you just made: a student who is deaf and keeps captions on out of habit, a commuter watching on a muted phone, and Google’s crawler, which treats a video as a black box with a title stuck to it. Captions and transcripts are how you serve all three — but they are not one feature you switch on. They are three different assets doing three different jobs, and creators routinely overpay for the wrong one.
This guide separates what captions do from what a transcript does, shows the tools and current prices worth paying for, and gives you a workflow that spends money only where it earns its keep.
Captions, subtitles, and transcripts are not the same asset
The words get used interchangeably, and that is where the waste starts. A caption is synchronized text shown over the video that includes speech and the non-speech audio a viewer needs — speaker changes, [laughter], a phone ringing. The W3C defines captions as “a text version of the speech and non-speech audio information needed to understand the content,” synchronized with the media (W3C WAI, Captions/Subtitles, accessed September 2026). A subtitle, in US and Canadian usage, assumes the viewer can hear and only transcribes or translates dialogue; the UK and much of the world call both “subtitles” (Wikipedia, Closed captioning, accessed September 2026). A transcript is the full text of the audio as a standalone document — no timing, not shown on the video, meant to be read.
Why the distinction matters for your budget: captions and subtitles live on the video and are consumed while watching; a transcript lives as text beside or under the lesson and is consumed by readers and search engines. Buy one thinking you got the other and you end up with either accessible video that has nothing to index, or an indexable page your muted-phone viewers never see.
The asset that actually moves SEO is the transcript
Here is the part most “add captions for SEO” advice gets backwards. A caption file — an .srt or .vtt uploaded to your video host — is timing plus text that the player reads; it is not reliably crawled as page content, and inside a gated course it sits behind a login where no crawler goes anyway. What search engines can read is a published transcript: real text on a public page. If you want a lesson’s content to be findable — a free preview lesson, a blog-adjacent tutorial — the move is to publish the transcript as text, not to hope the caption track gets indexed.
So the honest split is this: captions win accessibility and retention; a published transcript wins SEO. Muted autoplay is the retention case — a large share of mobile viewers start videos with the sound off, and captions are the only reason they keep watching instead of scrolling. Accessibility is both an audience you are otherwise excluding and, increasingly, a baseline expectation for anything sold to institutions or the public. For the wider accessibility picture beyond captions, see our guide to online course accessibility.
Tools, prices, and accuracy: what to actually pay for
You have three tiers: free auto-captions, subscription AI transcription, and pay-per-minute human captioning. The trap is defaulting to the cheapest (auto only) or the most expensive (human on everything). Here is the current landscape.
| Tool | Type | Price (Sep 2026) | Accuracy | Best for |
|---|---|---|---|---|
| YouTube auto-captions | Automatic | Free | ~85–95% (varies with audio) | A rough first draft you must correct |
| Otter.ai | AI transcription | Free 300 min/mo; Pro $16.99/mo | AI-level; expect to edit | Bulk transcripts of clear-audio lessons |
| Descript | AI + editor | Free 60 min/mo; Hobbyist from $16/mo | AI-level; easy in-app correction | Creators already editing their video/audio |
| Rev (AI) | AI subscription | Free 45 min/mo; $25.49–$47.99/mo | AI-level | High-volume self-serve captioning |
| Rev (human) | Human captions | From $1.99/min | 99%+ (vendor guarantee) | Flagship, evergreen, sales-driving lessons |
The pattern in that table is the whole strategy: accuracy you can trust costs either your time (correcting AI output) or your money (human captioning). Nothing gives you both for free.
Where captions and transcripts go on Teachable, LearnWorlds, and Kajabi
The mechanics are simpler than the marketing suggests, because most course platforms do not transcribe for you — they host the video and let you attach a caption file. The universal pattern, the same one YouTube documents for its own uploads, is to generate a caption file (.srt or .vtt), correct it, and upload it alongside the video so viewers can toggle it on (YouTube Help, Add subtitles & captions, accessed September 2026).
- LearnWorlds lets you upload subtitle files per video and carry multiple caption tracks, so it is the most caption-native of the three.
- Teachable and Kajabi host your video but expect you to bring the caption file — or to embed a host such as YouTube or Wistia that carries the track, in which case the captions travel with the embed.
For the transcript, do not bury it in the video player. Paste the corrected text into the lesson’s text area (every platform has one) under the video, and — for anything you want indexed — publish it on a public page too. That is the version a crawler and a skimming student both use.
The workflow that pays: auto-generate, correct, then reserve human captioning
Captioning every lesson by hand at $1.99/min is how a 40-lesson course quietly costs you thousands. Doing nothing costs you the muted viewer and the accessible audience. The workflow that respects both money and reach looks like this.
1. Auto-generate everything, correct what ships
Run every lesson through an AI tool, or YouTube’s auto-captions as a free first pass, then correct. Auto-captions are a floor, not a finish — names, jargon, and numbers are exactly what they get wrong, and those are exactly what your students screenshot and quote back to you.
2. Publish transcripts only where they can rank
Free preview lessons, tutorials, and anything public gets a published transcript. Paid, gated lessons do not need one for SEO — nobody is crawling behind your login — so save the effort and put it into the pages that can actually earn traffic.
3. Reserve human captioning for flagship lessons
Your sales-page demo, your most-watched module, the lesson your refunds hinge on — those earn 99%+ human captions. Everything else gets corrected AI. That is where the $1.99/min actually returns something instead of insuring a lesson twelve people finish.
Who should do this now — and who can wait
Do it now if you sell an evergreen catalog, have or want an international audience, know your videos get watched on phones, or sell to schools and companies where accessibility is a procurement checkbox. In all of those, captions and a transcript strategy pay for themselves in retention, reach, or a deal that would not close otherwise.
You can wait if you run small live cohorts where you are present to clarify in real time, your content is ephemeral (a one-off webinar you will not reuse), or your platform already bakes in accurate captions and you have no SEO ambition. Buying transcription tooling before you have a lesson worth watching twice is optimizing the wrong thing.
Caption what sells, transcribe what should rank
Captions and transcripts feel like the same chore, so creators treat them as one line item and get the economics wrong in both directions. Keep the jobs separate: captions for the people watching (accessibility and the muted majority), transcripts for the people and machines reading (SEO and skimmers). Auto-generate as your baseline, correct before anything ships, and spend real money — human captioning — only on the handful of lessons that carry your revenue. Do that and you are accessible, findable, and still solvent.
Once your lessons are watchable with the sound off, the next lever is the video itself: see our picks for course video editing software and the habits that lift course completion rates.
FAQ
Do YouTube’s automatic captions satisfy accessibility requirements?
Not on their own. Auto-captions typically land around 85–95% accuracy and miss names, terms, and numbers, while accessibility guidance expects captions that accurately convey speech and the relevant non-speech sounds. Treat auto-captions as a draft you correct, not a finished accessible track.
What is the difference between a caption and a transcript for SEO?
A caption file is timed text the video player reads, and it often sits behind a login; a transcript is standalone text on a page. Search engines index the published transcript, not the caption track, so the transcript is your SEO asset.
Should I pay for human captioning on every lesson?
No. Use corrected AI captions for most lessons and reserve human captioning (from about $1.99/min for 99%+ accuracy) for flagship, evergreen, or sales-driving lessons where an error costs you students or refunds.
