Captioning Your Podcast With A Transcript For Those With Hearing Loss

This week’s miniseries tackles the topic of podcasting for those with hearing loss… like me! Before I get to some real, practical advice for all podcasters, I need to pull out and climb up on the soapbox for a moment.

That soapbox issue: The decision around whether or not to include transcripts for a podcast episode is often dependent on the perceived SEO value of said transcript.

🤬 <-- Actual photo of me, right now, typing out those words.

First, putting up a shitty transcript of an episode has no (and perhaps negative) SEO value for a podcast’s website. SEO, like making great content, isn’t something that can be sprinkled on top. If you care nothing about accessibility and take only one thing from this post, please let it be that shitty transcripts will not help your website’s SEO.

Second -- and vastly more important -- is that's not why you make transcripts of your episodes available! 

Transcripts of podcast episodes are an accessibility feature, not an SEO benefit.

Captioning is added to the audio (and video) elements of your digital content so that people can consume those elements with their eyes instead of (or in some cases in addition to) their ears.

Comparatively speaking, my hearing loss is quite mild and easily (and expensively) corrected with discreet hearing aids. If we’ve met, I was probably wearing them, and probably you never noticed. 

Personally, I don't read transcripts. Because I don’t have to. I'm not deaf. But for people who are deaf or have significant hearing loss, they don’t have the option of listening. For them, reading a transcript is effectively their only option. And because you’re a smart podcaster who wants their content to reach the widest audience possible, you need to make allowances for those who have hearing loss.

My thinking on transcripts for podcasts has changed over time, mostly because the word “transcript” isn’t sufficient to describe the end-product. When making a transcript of this episode, should I strive for a literal word for word, accurate representation of the words that come out of my mouth? The obvious answer is yes. Yet, when I've tried that, I don't like the way I read. I always listen back to my episodes after they are produced, and I like how I sound. I sound like I know what I’m talking about. But when I use my eyes to read the exact words that have been accurately preserved as they came out of my mouth, I don’t like how I read. I read kind of like a moron. 

Making An Article Out Of A Transcription

Since mid-June 2019, I’ve taken a different approach. I use a service to generate an automated transcription of my edited episode. Then I take those words that came out of my head through my mouth... and I'll rewrite every single one of them so that the episode reads much, much better. My words become more coherent, and it's easy to scan back and forth through the text, something we really can’t do easily or efficiently in audio-only. 

I think my approach is better than posting a literal, 100% accurate transcript. During my process, I’m not removing meaningful content from my spoken words, though I may cut out repetitive statements that made sense in the audible form but become redundant and predictable in text. And I’ll add things, like this sentence and the last, when I see a need to add clarity to those thoughts that were in the audio. So in essence, this written creation actually has more info for those with hearing loss. Bonus! 

I think? I welcome input on my current thinking and implementation. Is taking the 10-minute audio monologue and turning it into this 1,267 word written representation the right thing to do? Not just an acceptable thing to do, but actively improving the text in an effort to make it better? Here’s a link to the raw transcript of this episode so you can see the difference. I welcome your thoughts, especially if you, too, have hearing loss. 

Why You Need An Article And A Transcription Of Your Podcast Episodes

Previously, my advice was to create a very detailed blog post/article for each episode and skip the transcription. But now, I’m thinking the advice I have been giving out for years is flat out wrong. Because I think, even if you’re doing the deep re-write of the actual transcript like I am, I think you need both. 

Let me be clear: Yes, you do need to write and publish a detailed blog post, article, or landing page for your episodes. No, a couple of paragraphs and a few bullet points of key topics is not sufficient. You need to spend the time to make a very good and purposeful detailed landing page for each of your episodes. Add some charts, graphs, and other visual components to make a piece of worthy digital content. Landing pages like that tend to get rewarded by our friends at Google. Low-effort blog posts? Not so much.

But that's probably not sufficient for servicing the needs of the hearing impaired. In most cases, the audio of your episode and the contents of this new landing page, another digital asset, are vastly different. If that’s the case for you, then you need to include a transcript of the actual audio episode. Yes, that means more work for you. Sorry?

Living With The Reality of Automated Transcriptions

Please remember that automated transcripting is less than perfect. “95% accuracy” sounds like a lot, but that means ~70 words or phrases would be wrong in this article. I cannot abide that. You shouldn’t either.

Automated transcriptions aren't great. In fact, they're bad and they're hard to read along with. So you're really not doing the people with hearing loss any favors by posting just a garbage automated transcription. (Link again to prove my point.) 

Yes, automated transcriptions are getting better every day. Perhaps someday the AI behind the best automatic transcription will be perfect. But we’re not there yet. Until the future arrives, you need to spend the money to actually make a readable full transcript of your episodes for people who have hearing loss. 

Transcripts make podcasted content accessible. You, the working podcaster, want people to enjoy your content. Therefore, you need to publish an excellent transcription of each of your episodes. Not an automated blob of text that cost you a dime a minute. If there are multiple speakers, each speaker needs to be identified, and their words accurately translated. Do the work.

Taking the free or cheap way out isn’t sufficient for those with hearing loss. In fact, it’s offensive. It shows you don’t care, something I also cannot abide. 

Accessibility matters. For people (like me) who have hearing loss, accessibility matters. It should matter to you as well.

Tomorrow I’m going to get into some technologies and tools to help you make your audio content more accessible. But before that, two things real quick:

• BuyMeACoffee.com/evoterra is running if you want to show your appreciation for this show. You're probably never going to buy the services my firm offers since you're a working podcaster, so buy me a virtual coffee instead.
• The new Flick group app is growing. People are having conversations with me and themselves, but it's just getting started. If you want to be one of the early members of this new Flick group (it's an app on your phone), come on in! It’s free to join, of course.

Tomorrow I’ll dive into some technology, plugins, and more to make sure that the audio you're making is fully accessible on Podcast Pontifications. 

Cheers!