Up Next

10 Critical Questions: Selecting a Transcription & Captioning Vendor


The following white paper aims to give a clearer insight and approach when choosing a transcription and closed captioning vendor.  Depending on your own specific needs such as, boosting SEO, or making your videos more searchable some generic questions will need to be asked before selecting a vendor.

Selecting a transcription and captioning vendor is not as easy you think it may sound. There are various elements which need to be considered.  Just like everything we purchase in life, we want to ensure that we are getting the best product for the service we require.

There are numerous variations in quality, pricing, and capabilities but in the flooded market of transcription and captioning its important to recognized that accuracy takes a much higher stance over pricing. This means that the cheapest options are not always the best options.

Below are 10 questions which are relevant and should be asked when selecting a vendor. Remember the questions can be tailored to suit your specific needs or the needs of your company, As everyone is different in terms of ideologies, wants, and requirements – the aim of the vendor is to make your captioning and transcription process easy and stress free.

What is your accuracy rate?

The number one question to ask is what is the accuracy rate of the vendor you are considering? As an online video platform it is no surprise that YouTube provide their own speech recognition captioning service named “automatic captions” – this is where they create captions specifically for YouTube videos. The accuracy rate for this specialized feature is 60-70%. The problem here is that 1 out of 3 words are wrong, this is significantly lower than the standard Verbit has set which is a 99% accuracy rate.  It is important to note here that when the speech is captioned incorrectly they are often outrageously wrong, ultimately if transcripts and captions are incorrect they do not serve the complete purpose in which they were intended.  Generally it is best to choose a vendor with a 99% accuracy rate.

Other accuracy aspects to take into consideration are, how accuracy affects SEO – this is specific where quotes have been taken out of text in order to rank higher in Google. Consistency – are your vendors consistent in maintaining accuracy despite the size or number of videos? And caption quality standards – do the processes comply with FCC, ADA, and the rehabilitation act requirements?

There are legal guidelines which need to be taken into consideration in terms of accuracy. For instance, the FCC recently released new quality standards for closed captioning. When focusing on accuracy the legal body states that, “Captions must match the spoken words in the dialogue, in their original language (English or Spanish), to the fullest extent possible.” Thus, in order to maintain the impact of the performance and portray the message which has been set across the captions must include non verbal sounds, such as different voice tones, fillers, background music/noises, and also be fit to adjust to ones accent. Furthermore, captions must also mirror the tone of the speaker in order to get an even more accurate perception of the message being conveyed.

SEO is still something many of use are trying to get our heads around without the use of apps and complex plugins. If SEO search is a dominant factor then accuracy will be even more important for you. Accuracy falls here from 99% to 98% – often due to speech recognition on names, places, object etc. Subsequently this means that using automatic speech recognition alone will likely register with Google as “automatically-generated gibberish” which in hindsight could have a negative impact on your SEO efforts

Ensure that your vendor has a consistent method of processing your files in regards to accuracy.They will need to deal with different types of content ,it is key that they will be able to maintain accuracy levels across all files.

What is the process of creating a transcript?

Don’t be afraid to get to know your vendor. Figure what they do, and how. This will also give you some insight the methods they take to maintain and ensure high levels of accuracy. Some ways in which transcripts and captions are created include the following:

Offshore transcription – this method tends to be cheaper and cost effective, however as mentioned above, “the cheapest options are not always the best”. Offshore transcriptions methods are not very good when it comes to accuracy and is not the best when handling English grammar and spelling. Would this work for foreign students who use English as a second language? Perhaps not….

Crowd sourcing – Although also cheap, the 2 main effects which come with crowd sourcing are consistent accuracy, and failure to comply with legal regulations. Be careful with this one!

Training – It’s important here to confirm if the vendor transcription and captioning are professionally trained and follow strict standards. It is also worth checking whether or not the vendor’s transcriptionists have domain expertise. This is particularly important for those dealing with academic content

Multi step process – This is the best and efficient way to produce accurate content. This is due to the fact that vendors are first leveraging the technology by putting the videos through speech recognition software rather than starting from scratch. Once this is done they then edit and checkfor spelling and grammar mistakes, and other general standards until the content is close to perfect.

How do you handle difficult content?

In the world of transcription and captioning there are so many different types content vendors will need to work with. So it’s important to understand how they deal with content on the more difficult spectrum such as physics, calculus, philosophy, law, and medicine. Anyone with videos that cover esoteric concepts or use of difficult language should ensure their vendor is able to handle the unique language.

Some ways in which vendors can handle difficult content includes, glossaries – provide this to your vendor so they are aware of the types of language to expect. Domain expertise – how do vendors select transcripts?  Give your vendor difficult content to transcribers who are most experienced in the field in order for them to draw from the domains and get the best result. Lastly, what is the editing process? After the video has been transcribed and filtered it is worth checking which editing process they use in order to ensure the end result is almost perfect.

What is your turnaround?

In today’s day and age its safe to say that human beings lack patience. With almost everything being at our fingertips gone are the days of having to wait upto 5 working days for a result. Que the turnaround time…

How long does it take your vendor to transcribe a piece of audio? While most vendors propose a same day turn around, this can vary depending on the size and quality of the media file. Generally there are three main turnaround options, these are standard turn around, selective turnaround – where it can be flexible dependant on budget and other aspects; and extended turnaround – where you get to prolong the process. However, it is key to note here captioning companies are notorious for missing deadlines, so be sure to ask about their track record and policies for remedying late files.

What video formats do you accept, and what caption formats do you provide?

Be sure to check which video format you will be using in order to ensure it is compatible with your chosen vendor.

It is also required that you think about where you are publishing your video. For example YouTube, DVD, Itunes etc – ensure that the vendor can output for your specified format. Good vendors do tend to support most formats.

Other questions to take into consideration include, does the vendor allow reformatting? And are you able to download your captions into different  formats?

Do you provide integration to alternate the work flow?

Integration can be useful addition to your process when using popular online video platforms such as YouTube. Its therefore beneficial to know whether or not your vendor offers this service.

Once integrations have gone through the video platform – the video is selected for captioning (based on the video platform or directly from your account). When this is done the captions are posted back to your videos.

What does the workflow look like?

Getting a good grasp of the vendors work flow is a good way of alternating your productivity. It will give you insight into what is being done to the videos and what needs to be done. Please not that if integrations are not offered by your vendor (where videos are only hosted in their servers) – you will need pay significant attention to the work flow.

Some other things to take into consideration include:

  • What are the upload options?
  • How are things such as turnaround, accuracy, transcript settings, and download preferences specified?
  • How do you download your caption files?
  • Is it easy to edit errors?
  • Once the captions have been downloaded are they deleted forever?
  • When switching platforms can captions and downloads be carried over?

What is the cost of captioning?

Price and costing is often dependant on size of files, turnaround, and other fees. The initial question should be how does the pricing model work. For example, does the vendor do rounding to the nearest minute? And is there a per file minimum.

Other fees – I’m sure we all hate the feeling of surprised hidden fees so be sure to confirm which extra costs you may entail. Some captioning vendors have one set fee, but may have added costs for extra services such as identification, certain caption formats, etc. Ensure you are aware of the total price, so there are no surprises.

A great way to save cost is to also find out if your vendor bulk discounts. For example, vendors may allow you to purchase a bulk number of hours at a discounted rate. So it’s worth checking this out!

Lastly, as mentioned earlier – price vs accuracy. Keep in mind that cost is often aligned with accuracy so you may want to shy away from cheapest of cheap vendors.

What are the capabilities of the account system?

Make sure you are aware and comfortable with the account settings, and find out if your vendor provides the service you require. Ensure that they are user friendly for yourself or any additional parties.

Billing and sub accounts are two features which many look for in the account system so make your vendor has these options available. Especially if you are in the education field. In particular when it comes to billing consider the following, how does invoicing work? Is all billing done electronically? Can you pay directly from within your account?  Is everything itemized?

Finally, consider how accessible the user interface actually is. For example, how are your files arranged? Are they accessible? How long are transcripts and captioning stored in your account? Can you download different formats later? Does the vendor offer tutorials? Is it evident where to download and upload? ? Can you easily set transcription, download, billing, and user settings?

Will I get reliable support?

Reliable support is valued in a company and it can be the make or break of a organisation. Ensure you can contact your vendor relatively easily and find out their main means of communication. Consider their response times, location, size of team, and how they can be contacted.

Up Next

Web Accessibility and Higher Education

For a moment, let’s consider accessibility and imagine that the year is 1990.

New Kids on the Block is churning out hit after hit, Home Alone is the No. 1 movie in America, and the Internet still exists as a hard-to-explain phenomenon (yes, a strange year indeed).

And yet, beyond the pop culture references and fun facts, 1990 also marks a watershed moment for the modern civil rights movement. It is the year that President George H. W. Bush signed into law the Americans with Disabilities Act, or “ADA.” The ADA, in its current form, exists to prevent discrimination based on disability within both the public and private sectors. These sectors include (among other areas) transportation, employment, and – most importantly – education.

Reflecting the necessity of providing a “full and equal” education to persons with disabilities, enforcement of the 1990 Act has traditionally focused on physical, on-campus spaces – i.e., access ramps, lavatory facilities, classroom accommodations, etc. However, with the rise of the internet and the rapid evolution of modern technology, America’s institutions of higher education now face a new and formidable ADA compliance challenge – the university website.

Under the wording of Title II and Title III of the Act, most universities (both public and private) are required to provide students with equal access to information and communication technologies. In 2018, this means not only providing all students with an open and easily accessible website, but one with effective information communication as well (i.e., accurate transcriptions and precise closed captioning)

Transcription is a hot topic, but not a new one; we discuss this more in our previous blog post. As we will soon see below, the consequences for failing to comply with the ADA can be both severe and long-lasting.

Why Is Website Accessibility Important?

It’s easy to assume that an open, accessible, and adequately captioned university website is a foregone conclusion in 2018. And yet, proper website captioning (or a lack thereof) is quickly building steam as one of the most critical ADA compliance issues of the digital age.

How much steam?

Well, according to a list compiled by the University of Minnesota, more than 30 colleges and universities have been sued or have had complaints lodged against them for failing to provide disabled students with sufficient access to information. Harvard, MIT, and UC Berkeley, prestigious and internationally acclaimed universities, have all found themselves out of compliance with closed captioning requirements, and in turn, as parties to federal complaints and costly lawsuits.

ADA lawsuits graph

The Harvard and MIT lawsuits, filed in 2015 by the National Association of the Deaf, are seeking a permanent injunction requiring the universities to provide closed captioning and text versions of all online materials. The lawsuits address the plethora of “inaccurately or unintelligibly captioned” online content, “making it inaccessible for individuals who are deaf or hard of hearing.” In 2016, the District Court of Massachusetts denied each university’s motion to dismiss and ruled that the case should move forward. As of writing, this costly case is still pending with no end in sight.

While the Harvard and MIT cases provide us with illuminating examples of the potential legal consequences for failing to provide online closed captioning to students, the case involving UC Berkeley highlights another potential accessibility pitfall altogether: non-student access to publicly available material.

In October 2014, the National Association of the Deaf, together with the Civil Rights Education and Enforcement Center, filed a complaint with the U.S. Department of Justice (“DOJ”) alleging that the deaf and hard-of-hearing could not properly access UC Berkeley’s no cost audio and video content. In a ten-page letter sent to university officials, America’s premier public university was admonished by the DOJ for (among other offenses) the use of improper video formats, poor color contrasts, and a significant shortage of closed captioning transcriptions resulting in the content of the videos being “inaccessible to people with hearing disabilities.”

UC Berkeley, rather than attempt to comply with the costly recommendations of the DOJ, made the remarkable decision to remove public access to thousands of videos and assorted materials. As reported by Inside Higher Ed, UC Berkeley choose to remove more than 20,000 audio and video files from its publicly accessible platforms rather than pay the costs of compliance. Both UC Berkeley faculty members and other experts/scholars have expressed concern regarding the administration’s justification for its decision, as well as what appears to be an apportionment of blame towards disabled persons.

While the DOJ did not make findings regarding UC Berkeley’s treatment of its student body, the moral of the story is this: even when avoiding the legal consequences and financial costs of web accessibility compliance, universities can still potentially damage their carefully cultivated reputations by not complying.

When in doubt – comply, comply, comply.

How Should Universities Approach Compliance?

There are many ways universities can ensure compliance with ADA web accessibility regulations. The following are just a few of the many best practices as recommended by the World Wide Web Consortium (WCAG 2.0):

Informative descriptions of hyperlinks
Easy to understand menus and navigations tools
The use of alternative text for images
Clear and consistent headings
Avoiding reliance on color or sound for navigation
While all of these recommendations go a long way in ensuring ADA compliance, perhaps the most significant issue universities face when it comes to web accessibility (as illustrated in the cases above) is the matter of closed captioning. As education continues to evolve in the digital age, most colleges and universities will soon have thousands upon thousands of podcasts, videos, and other similar types of media content – all requiring proper closed captioning and transcriptions. And although the sheer volume might appear overwhelming, it is important to remember that proper captioning is an incremental process. It will not happen overnight.

Institutions implementing closed captioning

A revealing survey from Oregon State University recently highlighted both the anxieties and narrow motivations of higher education when addressing closed captioning. The study indicated that most institutions implement closed captioning for the sole purpose of complying with the law. And yet, it should be recognized that there are many additional compliance benefits to consider. Not only are visitors more likely to return to a website that has closed captioning, but it is no coincidence that accessible websites also happen to be the most useful websites. ADA compliance, although it applies specifically to persons with disabilities, will likely result in a website that is clearer and easier to navigate for everyone. Website traffic also stands to increase, as an easily accessible website should be easier to find in search results.

It is critical to remember that proper closed captioning begins with a shift in institutional mindset. It continues with intense preparation and planning, and it finishes with the right application of AI technology and human knowledge.

It is a process to help you better serve your community, and it is one in which we at Verbit can help.

Back To Top