In a nutshell: Podcast Transcription is essential to good PR and podcast accessibility. This article covers three means of getting it done.
- You can do it yourself, with some simple tools.
- You can hire a freelancer.
- You can try a service. This article will also discuss the pros and cons, whys and hows, of a few different transcription services.
Subscribe to the Podcasting Advice show on your listening app of choice.
We think of podcasting as a purely aural medium, but we ignore podcast transcription at our peril (not to sound like a character from I, Claudius). In all seriousness, if your podcast doesn't have transcripts, that's an awful lot of audience you're missing.
“Wait a minute!” you say. “Transcripts are more time and money. How much do you expect me to invest in this podcasting thing? What's in it for me?”
We've already talked about why you should use transcripts. It bears repeating, however:
- 11 million people in the UK and 10 million people in the US are d/Deaf or hard of hearing
- Many folks have auditory processing disorders
- Transcripts of your podcast episodes can boost your search engine optimization
- Transcripts are another avenue for more people to understand and engage with your content.
In this article, we're not going to talk about why you want to invest the additional work or money into podcast transcription. We're going to talk about different methods or companies that you can use to get your podcasts transcribed, and in front of your audience.
DIY Methods for Podcast Transcription
If your podcast is a scripted audio drama, you can certainly use that script. You can also sit down and type out your podcast's recording, word for word. If you're already a professional stenographer, this won't bother you too much. However, not many people have that skill set or patience.
Record. Edit. Publish. All in one Tool.
Alitu records calls, solo segments, cleans up your audio, adds music & transitions, helps you edit & publishes right to your host.
Dictation Software Hack
- Open up Chrome (you have to use Google's Chrome browser, otherwise the tool isn't available)
- Go to Google Docs and open a fresh document
- Click on Tools and select Voice Typing from the drop-down menu. A nifty little icon of a microphone will appear. At the top of the microphone, select your language.
- Open up your DAW and start playing the sound file you want transcribed.
- Click the little microphone icon. It'll change colors. As the microphone picks up the dialogue, the text will appear on screen.
When you try this, make sure that the audio is cued up at the point where you want it transcribed. if you let it run through a minute of intro music or so, the voice typing tool won't recognize the music as anything useful, and will shut itself off.
Google's Voice Typing was pretty good at recognizing what I said when I tried it. To test its accuracy, I selected UK English, US English, and Nigerian English, to see if there would be differences in the resulting text. There wasn't.
You'll end up with pages and pages of unpunctuated text. It's very stream-of-consciousness. You will have to go through it and format it, put in punctuation and check for errors. However, it saves you a lot of time over typing the audio word for word.
There's a quality control issue here, as well. You may notice as you're making this podcast transcript, that you notice the dialogue as an outsider. Transcribing the podcast yourself will help you see the details of what you talk about, and how.
Another DIY method is to use your audio to make a video, upload it to YouTube, and use its closed captioning to provide a transcript. I can't vouch for the accuracy of YouTube's closed captioning. Joe Rogan uses this exclusively, and it works for him. Your mileage may vary.
Headliner is a free app which can be used to make audiograms (and it's fun to use). It's 100% worth your time to try it out, and see how its captioning system fits your podcast.
Think about your SEO for a second. Google's Product Manager for Google, Brad Ellis, spoke about automatic captioning and SEO at the Streaming Media Conference in 2016. He said that when a YouTube user uploads their own captions, Google does index them. But, because the automatic transcriptions are error-prone, they are not indexed. Granted, this was four years ago, and the technology may have changed. For now, if your SEO is your most important reason for having transcripts, it's best not to rely on YouTube's automatic captioning only.
Many people consume their podcasts via YouTube. It's not a bad idea to make them available there with captions. But it shouldn't be your only way of publishing those transcripts.
Outsourcing Podcast Transcription
If you don't want to do the transcription yourself, you have a few options. They are:
- Humans (having someone else do it for you)
- Artificial Intelligence (a software does it)
- A combination of both
This is the most accurate version of podcast transcription that you can use. Humans can understand context and spoken sound in a way that machines can't. If your podcast uses the sentence, “Let's eat Grandma,” a human would know where to put the comma, while a machine might not. It's also not cheap. Some transcribers charge by how long it takes them to do the work, others by how much audio you supply. In most cases, it's between $1-$2 per minute of transcribed audio. If you look at a site like Upwork or fiverr.com, you can find a fair number of freelance transcribers.
GoTranscript offers “100% Human-Generated Transcription Services,” and offer a minimum of 6-hour turnaround time. Their prices vary by how soon you need your podcast transcript finished, and how long your audio files are. Expect to spend around 80 cents to $2 per minute.
If you have a loyal fan or close friend who loves your podcast, you may be able to get them to volunteer to transcribe your podcast in a skill swap. Ask them if they would transcribe an episode or two of your podcast in exchange for helping them move house.
If you've ever asked Siri to send a text message for you, or if you've asked Google to look something up, you've used something like one of these services.
Trint offers monthly or annual subscription packages ($60-$85), with a 7-day free trial. They range by the amount of files you can upload per month, and how many users you can have on the account. However, since a file can be any duration, the starter package would fit most indie podcasters, if you organize your audio files well. Trint also has an iPhone app.
Temi is very similar, with one price tier: 25 cents a minute. Their free trial provides one transcript, of a 45 minute audio file. Their free demo not only shows you how it transcribes a news piece with multiple speakers and accents, but also will let you practice editing the piece to test the interface. This free demo is 100% worth your time to look at, because it can show you good transcript layout.
Descript offers a free trial for the first 3 hours of speech uploaded. After that, the service is $14-18 a month. They say that they can turn around a 60-minute sound file in about three minutes.
Sonix is another quick and inexpensive podcast transcription service. The Standard package is $10 per hour of uploaded audio. The Premium tier is $5 per hour and $22 per user, per month. Sonix provides transcription in multiple languages, even at the Standard level. Their free trial includes 30 minutes of transcription service.
Artificial and Human Intelligence
These services use a combination of AI transcription, with a human checking for accuracy and consistency.
Scribie offers a pay-as-you-go system, either fully automated for ten cents a minute, or eighty cents a minute for automated transcription with a human editor. They add on an extra fifty cents if your audio is poor, noisy (speaker isn't distinct from the background), or for “accented speakers.” They say that it's fifty cents a minute for American speakers. This makes me want to send them a sound file in an accent so Philly, it sounds like I live underneath the Market Frankford El. An integrated editor lets you make changes if necessary.
Rev.com might be the most well-known podcast transcription service; it was included on Forbes' AI 50 list. Its automated transcription costs 25 cents a minute, with a 5 minute turnaround time. Their human transcription service is $1.25 a minute, with a 12-hour turnaround time. They can even create foreign subtitles for video, for $3-$7 per minute. Some podcast hosting services (such as Castos) include transcriptions using Rev.com for an additional fee.
Podcast Transcription: Summary
Your podcast transcript can build and sustain your audience. It can bolster your SEO, and make your content more meaningful to others. Remember, audio transcription services in general rely on people who need work-from-home jobs, such as disabled people or caregivers. Not only are they working on your podcast transcription, they can be working on medical lectures, court documents, emergency calls, things that are high-stakes, taxing, boring, or disturbing. The compensation for transcription workers is a controversial topic, one worth discussing on its own. Take this under consideration, when you're weighing price and turnaround time. It's probably the most ethical option to hire an individual freelancer, or barter a skills trade for podcast transcription with a trusted friend. As you choose how to release your content, you may want to consider the relationships you make, and businesses you support, along the way.
Need More Help?
Good podcast transcripts aren't the only element of podcast promotion. The Podcast Academy has courses and resources to help you with everything from mic technique to unlocking the secrets of DAW. Plus, the weekly Live Q&As provide an opportunity for you to get specific assistance on any aspect of podcast creation. Join us now.