Once upon a time, copywriters clicked away at typewriters and breaking news came in on teletype machines that went clackity-clack.
We’ve come a long way since then. Today, copywriters have Google Drive file sharing and real-time editing. Marketers have an entire suite of automated marketing technology. And journalists can instantly share whatever they’re reporting on via real-time streaming and social media.
The explosion of content-related technology in the 21st century has made it easier than ever before to gather, report, distribute, and share news and content. Yet one part of the modern workflow—transcription—remains largely unchanged.
It seems odd that one of the most mundane tasks in the newsroom production process—getting interviews on paper—should still be so unwieldy. But when you think about every step of the traditional transcription process, it kind of made sense:
Step 1: You would record an interview with someone who has important things to say, things that your audience really needs to hear.
Step 2: You’d playback the recording _over and over again: PLAY. STOP. TYPE. REWIND. PLAY. STOP. TYPE. REPEAT… As this was happening you would write out the transcript or in years past you would use a typewriter, if you’re into that sort of thing.
Fast forward a generation: the typewriter is now a computer, but the transcription process hasn’t changed much.
Problem: There has been no reliable and time-efficient way to get spoken words to magically appear on paper.
Solution: Enter AI.
In the past few years, advances in AI automation have unlocked previously unthinkable superpowers for the everyday journalist or marketer. Three of the most important new technologies include:
All three of these technologies will completely transform the business world as we know it (they’re already beginning to). But only one of these technologies is already at a fairly mature stage — automatic transcription.
If you work in media or marketing, you probably have some idea of how much audio-video content your organization produces over the course of a year, month, or week.
Well, here’s the thing. Whenever you write a new webpage or blog post, Google’s “spiders” crawl over that new text content and index it into their search engine. That way, whenever someone types in a relevant keyword, your page may show up. However, Google (and Bing, and everyone else) still can’t do this for video or audio content.
So if you have an hour-long video interview posted on a webpage, Google can find that video mostly because there’s associated metadata and a text headline. But finding specific parts of that interview, based on specific spoken keywords? Not yet possible.
Yet we’re all publishing more and more audio-video content. According to Cisco, 80% of all internet traffic will by video by 2018. The tragedy is that, currently, your internal production teams need to transcribe all of that audio-video content into text so that Google and friends can crawl them. Most people do this by hand. Naturally, it takes a ton of time.
If the vast majority of content will be video by 2018, then we’re clearly doing something wrong. We can blame the search engines for not advancing quickly enough, but we can also root through our toolbox to see how we can make the audio-video production process more efficient.
For now, we all still need transcripts for our audio-video content. This is the only way search engines can crawl that content effectively and deliver it to the right person at the right time.
With that in mind, it’s absolutely imperative that media and marketing organizations perfect their transcribing. This will help them save more time, more money, and get the most of their hard-earned media.
If you need some help automating your manual transcription process, reach out to us at firstname.lastname@example.org.