Insider Tip Six: Finding the Right Transcriptionist for You

You wouldn't think it would be hard to find a good transcriptionist today with the economy the way it is and so many people desperately looking for work. However, there's a lot more to finding a transcriptionist than just Google-ing for one.

First of all, there's the matter of training and experience. Not every transcriptionist has had the right kind of training to be able to do your job correctly and quickly. At the very least, your transcriptionist should have a Bachelor's degree as well as some type of special training in English grammar and spelling or medical transcription training which teaches all of the above and more. Grammar is not taught well in many schools these days, so don't naturally assume your potential transcriptionist can spell, type, or even speak English correctly. Ask for a resume or CV, and check out the person's references.

Then there's the matter of speed. Not everyone types at the same speed or has the same error free result that you would like. I personally only type 75 corrected words per minute (CWPM removes any typos before the speed is calculated). That's not as fast as many transcriptionists, but I am also keenly aware of errors and typos and try to avoid them the very first time around if possible (hence, the slightly slower speed). Just because someone can hit the keys at 100 words per minute doesn't mean you'll be able to read what they typed when they're done! Whether or not they proofread their documents is also crucial and something to ask about when looking for a good transcriptionist.

Along with the typing speed, you need to consider turn around time. It costs more to get your documents in 12-24 hours. So perhaps you could plan ahead on your audio or video files and then allow your transcriptionist a little more time to work on them (like 2-3 days, or 4-6 days). That would be significantly cheaper for you.

The last consideration when choosing a transcriptionist is, are they agreeable or disagreeable? This is very important consideration because, believe you me, there are some very disagreeable transcriptionists out there! They can be some of the most grouchy and touchy people around. Part of the reason for that is transcription is an intense job if you do it full-time. Put yourself in their shoes. Sitting at a desk in some corner typing as fast as your sore fingers and carpal tunnel syndrome can go for 10 or 12 hours a day does not a happy person make. That's why I would suggest that you conduct some correspondence with your transcriptionist before hiring them and try to ferret out their personality traits, true feelings, and work ethic. It could save you a lot of trouble in the long run. One of the reasons I only do transcription part-time is because I want to continue to be a happy transcriptionist with smiling satisfied customers.

Insider Tip Five: How to Prepare Your Audio/Video for a Transcriptionist

Let's say you've conducted a webinar, teleseminar, call-in program, or conference of some sort, and you're now ready to have your audio or video transcribed into a document to post on your website. There is an important step that will help your transcriptionist tremendously when s/he transcribes your document for you. This step is to figure out how many people in total are speaking on the audio or video. If there are a lot of speakers or people asking questions, this may be a bit difficult for you to figure out. However, it's definitely worth your time. It's also worth it to preserve the sanity of your transcriptionist!

Once you've figured out how many people are speaking, you'll need to clearly identify each caller, spelling their first and last names correctly on a list in the order in which they're speaking. These people will need to be mentioned by name on the transcript of your webinar/teleseminar, and they will want their names spelled correctly in the document (who doesn't?!). This is a huge help to your transcriptionist, and it will prevent them from listing people simply as "male speaker" and "female speaker" which gets very confusing, especially in a group setting where there are multiple males and females speaking.

Another thing you will need to do, especially for video conferences, once you have the number of people speaking and the list of their names is to identify the person in some way. You can mention the color of their hair, the length of their hair, their lack of hair (if it's a guy), the color and type of their clothing, or possibly even their accent ("Australian accent" for instance). This will help your transcriptionist more than you could ever know, and the person speaking or asking a question can even be picked out with their back to the video camera (which is often the case with conference videos anyway).

By following Insider Tip #5, you will make a fast friend of your transcriptionist, which leads us right into Insider Tip #6.

Mispredicted Words, Mispredicted Futures

The accuracy of computer speech recognition flat-lined in 2001, before reaching human levels. The funding plug was pulled, but no funeral, no text-to-speech eulogy followed. Words never meant very much to computers—which made them ten times more error-prone than humans. Humans expected that computer understanding of language would lead to artificially intelligent machines, inevitably and quickly. But the mispredicted words of speech recognition have rewritten that narrative. We just haven’t recognized it yet.

After a long gestation period in academia, speech recognition bore twins in 1982: the suggestively-named Kurzweil Applied Intelligence and sibling rival Dragon Systems. Kurzweil’s software, by age three, could understand all of a thousand words—but only when spoken one painstakingly-articulated word at a time. Two years later, in 1987, the computer’s lexicon reached 20,000 words, entering the realm of human vocabularies which range from 10,000 to 150,000 words. But recognition accuracy was horrific: 90% wrong in 1993. Another two years, however, and the error rate pushed below 50%. More importantly, Dragon Systems unveiled its Naturally Speaking software in 1997 which recognized normal human speech. Years of talking to the computer like a speech therapist seemingly paid off.

However, the core language machinery that crushed sounds into words actually dated to the 1950s and ‘60s and had not changed. Progress mainly came from freakishly faster computers and a burgeoning profusion of digital text.

Speech recognizers make educated guesses at what is being said. They play the odds. For example, the phrase “serve as the inspiration,” is ten times more likely than “serve as the installation,” which sounds similar. Such statistical models become more precise given more data. Helpfully, the digital word supply leapt from essentially zero to about a million words in the 1980s when a body of literary text called the Brown Corpus became available. Millions turned to billions as the Internet grew in the 1990s. Inevitably, Google published a trillion-word corpus in 2006. Speech recognition accuracy, borne aloft by exponential trends in text and transistors, rose skyward. But it couldn’t reach human heights.

Source: National Institute of Standards and Technology Benchmark Test History (borrowed from: http://robertfortner.posterous.com/the-unrecognized-death-of-speech-recognition)

Insider Tip Four: How Long to Make Your Audio/Video/Articles

Quick Summary: Don't record more than 30 minutes at a time--it will be too cumbersome a document to read otherwise, unless you plan to create several chapters in an e-book or it is part of a larger body of work.

As I have transcribed articles for various Internet marketing experts, real estate big wigs, and teleseminar gurus, I've noticed something quite interesting. They lose me after about 30 minutes of their "expert" audio/video. In spite of the fact that you can cram a whole lot more useful information in a one hour or a 1 1/2-hour audio file, you're just going to lose people along the way. They'll be dropping off like melons off a cart on a bumpy road. That is especially true when your audio/video is transcribed into a document.

A 30-minute audio recording will render about a 13-page document. From my perspective, I can tolerate reading 13 pages at one sitting. However, your average reader today who often reads on an 8th-grade level or below will take longer to read 13 pages. People who speak and read English as a foreign language will even take longer than that. Others are used to short synopses or summaries on various Internet sites and blogs and don't have the patience for longer articles. As I mentioned in Insider Tip Three, considering your audience is crucial, so you need to remember these aforementioned groups when you're planning your audio/video or article.

If your recording is an hour or longer, it will produce about 26-30 pages once it has been transcribed. That's a fairly long document to weed through for most people. I would say unless you plan to make your document into an e-book or a printed book with chapters of some sort, you're better off sticking to the 30-minute rule.

There are always exceptions to the 30-minute rule, for instance, sermons, which are often a bit longer. That's to be expected. However, if you're writing for the Internet, keep it short and sweet. Use your words carefully and succinctly--they will cost you in the end.