So, you want to use transcription services, and you know you'll need to partner with a great team and process in order to get that done. But are you ready to upload that file yet? This guide will take you through all of the things you need to ensure that you get the transcription services fit for your requirements — from picking the right file formats and the difference between software-led and high-quality transcription services, to the questions you should ask your transcription partner.
Transcription services deliver speech to text. Most commonly, this is through digital audio files. But, video to text services are common. A few services provide specialist in-field notetakers or stenographers that will come to you and transcribe events as they happen.
On the face of it, transcription services are simple and straightforward. You find the best rate online, create an account, send in your recording and download a text transcript. There are, however, important distinctions between the types of transcripts you will receive and why certain services cost more than others.
No, they are not. The most obvious difference is quality or usefulness and cost. This can range from automated transcription software services with free ‘basic’ options, to human-based services with fees over £2 per minute of recorded audio. Fundamentally, these are different products that represent two distinct categories within transcription services.
Human transcription services produce accurate transcripts that you can count on — automated speech to text software does not. For an ‘ideal’ and high-quality recording, the difference can be slight. However, independent comparisons demonstrate that audio to text software transcription services
Human transcription services consistently hit accuracy rates above 99%, even when dealing with poor audio quality, multiple speakers and complex language. If you are willing to pay for it, even better outcomes can be guaranteed. Most human transcription services will remove ‘ums’, false starts and stutters to create a smoother reading experience without damaging meaning. Using transcription services allows for customised outcomes and the transcription of audio files that would be indecipherable for software.
Audio quality, multiple speakers, the speed at which individuals are speaking, accents, background noise and the complexity of language all impact automated transcription quality. For human transcribers, these complications will slow down the process and can drive up costs, but a quality outcome will be delivered. For speech to text software, one or more of these complications can create nearly unintelligible outcomes. Fundamentally, you will need to review and edit any transcription made using an automated service.
Transcripts take many forms. With automated transcription services, it is always possible that your transcript will look like gibberish. All software transcription platforms attempt to deliver a verbatim account of your audio file (including ‘ums’, ‘errs’ and false starts), and attempt to distinguish between different speakers. Although most automated transcription services claim to be able to distinguish between speakers, this is one area where they struggle — along with deciphering a clear transcript when there is background noise, poor audio quality or certain accents.
Human transcription services deliver error-free transcripts with different speakers accurately distinguished, even with poor audio quality. They also deliver options.
There are three main categories of transcript: word for word, summary/notes and verbatim. These offer different levels of detail vs. readability at a number of different price points — allowing you to tailor results to suit your needs.
When your professional reputation is on the line, it is important to think about who is transcribing your audio for three reasons: quality, accountability and security.
The content that businesses and individuals want
A lot of companies will outsource transcriptions abroad to bring down costs. If you find human-based transcription fees starting at less than £0.75 ($1) per recorded minute, you can pretty much guarantee that offshoring is how those fees are achieved. From a cost perspective, this is great. It is not optimal from a security or quality standpoint.
Accents and local colloquialisms present challenges. Meaning can be missed and less accurate results delivered. This is particularly important when relying on the industry standard of ‘intelligent verbatim’ and even more important if you get a ‘summary’ transcription.
The use of double negatives, technically incorrect conjugations and slang are all regionally relevant. Industry jargon is professionally relative. It is important that the person transcribing your audio understands the meaning in order to make an accurate transcript. Look for transcription services that offer transcriptionists in your area and those that offer transcriptionists with relevant professional experience.
When engaging with offshore transcription services, it is possible that you will suffer extra fees because of accents, removing the cost-saving advantage of using offshore labour. Check to see if the transcription service reserves the right to increase their fees based on ‘difficult’ audio, look into how that is defined.
By engaging with a transcription service (rather than transcribing the material yourself), you lose a degree of control. It might be wise to get your transcription partner to sign an NDA (non-disclosure agreement) — most are willing to do so. However, you need to think about their ability to enforce that agreement and your ability to pursue violations.
If your files are sent abroad, it becomes more difficult to prosecute any breaches in security — a reality that increases the likelihood that such breaches might occur. Fundamentally, your transcription service partner will have degraded their ability to guarantee compliance by offshoring their labour.
If you are concerned about the security of your data, it is important to think about who is transcribing it. Try to keep data in regions with high ethical business standards and robust legal systems. You will often find it easier to prosecute violations in your home country.
It is important to pay fair wages. If you work with a company that keeps transcriptions local, you know that you are not engaging in exploitation. Although companies that employ offshore labour charge less, they also often keep a more significant percentage of that fee. Think about the types of working practices you are supporting.
If you pay people a fair wage, they will do a better job. If you want someone creating your transcripts who feels valued and is interested in their job, you are going to have to find a company that keeps its employees in-country. You are also probably going to have to pay a premium. The advantage is that you will get a better product that someone spent more time and effort creating.
Why UK Transcription Services Are Better
World leading European data protection regulations (GDPR) make UK transcription services a particularly safe option for exporting English transcriptions — delivering greater security and accountability. This also helps companies subject to the GDPR feel confident that their transcription partner is following all of the relevant regulations.
The UK has a long local history of pursuing privacy violation and a robust legal system that ensures high ethical business standards and a low probability of security breaches. Culturally, the UK is accommodating to accents. The wide variation across the country, large multicultural populations and proximity to European accents has created a population accustom to parsing the fine details of different English dialects and guarantees the ability to capture the nuance of your recording. The weak pound makes it cheaper than you think!
There is a lot of variability in the cost of transcription services and an equal number of variables in outcomes. Automated and human transcription services sit in somewhat different product categories.
Automated transcription services are relatively straightforward with pricing, although different companies use different models. Some operate on subscriptions, others have additional or standalone per minute fees. With the exception of platforms like Otter that have relatively generous free packages, you can expect to pay around £0.08 ($0.10) per minute of recorded audio time. Subscription models range between £12 to £7 per month.
When subscriptions are an additional fee, they often bring with them premium services such as timestamping that would otherwise cost extra. When subscriptions are the sole cost, they are often limited by time restrictions per month, as are free services. In all cases, the price differences in automated services are mostly driven by the quality of the interface — export options, online editing, mobile syncing and support. The quality of the transcriptions are all relatively similar, and all equally volatile depending on the quality and nature of the audio.
Time caps and fees can create pricing complications if you use software transcription services a lot. However, the options and quality control guarantees that fluctuate costs with human transcription services are not present. The real challenge with automated services is determining how compatible your audio quality is with a quality outcome. It is worth exploring the free trials on offer if you think automated transcription services will suffice for your transcription needs.
Human transcription services offer more options than transcription software platforms, creating more complex pricing models. However, they deliver certainty and control over what you get. Many pricing variables are linked to quality control. If you need quality and options around the detail captured in your transcription, then human transcription services are how you get that — no matter what it costs. This is particularly true when you require high levels of accuracy, have complex requirements or poor quality audio files.
Starting prices vary based on the type of transcription you want. Strict verbatim transcripts cost more, summaries cost less. Intelligent verbatim is seen as the industry standard. In most cases, transcription services will level fees for poor audio quality, multiple speakers, timestamps and accelerated turnaround times.
Human transcription services start at around £0.40 per minute of recorded audio. But, rates below £0.75 per minute probably come with a caveat — turnaround times in excess of a week, offshore workers or easily triggered and steep price increases based on audio quality.
To get a detailed and quality transcript of complex audio in under 24 hours, you should look at prices less than £1.50 per minute of recorded audio as a good deal. If using a company for the first time, or concerned about a tight budget, it is a good idea to call and get a direct quote regarding your specifics. The benefit of human transcription services, however, is that although it can be hard to tell exactly how much it will cost, you know exactly what you are going to get — a high-quality transcript of your audio file. This is effectively the opposite of what software transcription services deliver.
Turnaround time for transcription services depends on the type of service and your budget. Automated transcription services are nearly instantaneous. Most platforms can turnaround an hour of audio in a few seconds. However, you are likely to need to spend a significant amount of time cleaning up that file yourself to meet accuracy standards or even make the transcript legible. Automated transcription services struggle with background noise, poor audio quality, accents and multiple speakers.
You can expect to wait about a day for the transcription of an hour of audio using human transcription services. Some services offer lower prices in exchange for slower turnarounds (sometimes more than a week), while others will provide fast turnarounds in exchange for a premium — delivering transcripts in as little as 12 hours.
For custom jobs, you might be able to get even faster turnaround times. However, you eventually hit the limits of human capabilities. Skilled transcribers need about one hour per 15 minutes of recorded audio to transcribe a high-quality document — maybe more if the audio quality is poor. Some transcription services offer ‘notes’ options where a typist transcribes live, without pausing the recording. This type of service can be done faster, and normally costs less per minute, but delivers a ‘rougher’ outcome. As a base rule, per hour of audio, anything less than 24 hours is likely to cost you a premium fee, and anything less than 4 hours might be impossible.
Most human transcription services guarantee quality control under all realistic circumstances, although poor audio quality may incur fees. But, there are limits to reason. If the audio quality is too poor, some segments of the transcript might be lost. When it comes to software transcription services, audio quality will significantly impact the output. High audio quality and clear speech without accents are absolute necessities when using automated transcription services.
Yes, they probably will. However, accents are relative. This is one reason to seek out local transcription services. Most automatic speech recognition (ASR) software only excels with neutral American accents. Human services have a wider range of capabilities.
The main thing customers can do to improve the transcription process is to ensure high-quality audio. Get a dedicated audio recorder (such as a dictaphone or microphone attachment for your phone), make recordings in-person, in a quiet space and (when possible) get individuals to speak slowly, clearly and avoid speaking over each other.
If you can do all of that, you might be able to effectively use automated transcription software. That choice will limit your other options around ‘intelligent verbatim’ and summary transcripts offered by human transcription services. But, it will cost less. Under all circumstances, however, clean audio files will be quicker to turn around and will cost less to transcribe.
If your recording is of a complex subject and includes a lot of industry-specific or technical jargon, some transcription services will allow you to submit a vocabulary list/glossary. This doesn’t have to include everything. However, simply writing down any acronyms, brands or other jargon likely to come up in the recording will ensure that those words and phrases are transcribed correctly.
Most human transcription services come with proofreading included. Your choice is more about subtracting proofreading from the fee. Sometimes, however, you can pay extra for an ‘expert’ proofreading that will improve a 99% accuracy rate to a 100% guarantee.
The importance of proofreading in quality transcription services depends on what you want. If you need something error-free, then proofreading is critical. If you just need something that is legible and conveys the meaning of the recording, some services will offer you a cheaper alternative. When it comes to automated transcription software, you are going to have to do the proofreading yourself and take any actions to remedy the inaccuracies.
Confidentiality and security are essential to many people and businesses seeking transcription services. Google reviews to make sure the company is legitimate. Only use websites that have an ‘https’, rather than an older ‘http’ address. ‘Https’ encrypts communication between the website and your browser. This is an important thing to watch out for when entering any details online.
Make sure that transcription services use SSL encrypted login portals for sharing data and store information in encrypted formats across browsers and servers. Encrypted data storage is essential to ensuring that you are protected against a breach. This might also be vital to keeping you in compliance with data protection acts. This is particularly true for European companies that are subject to the GDPR.
The two big indications of a safe data environment are ISO 27001 & 9001 certifications. Be wary of engaging with transcription services that don’t possess these accreditations. If concerned about GDPR, it can help to contract with transcription partners based within countries covered by the GDPR. For companies subject to European jurisdiction, it is essential to make sure that your transcription partner follows all GDPR requirements, not just those regarding data protection.
Most transcription services are willing to sign an NDA (non-disclosure agreement). This is a good idea. If you are concerned about security, however, you should think specifically about who is transcribing your data. Secure login portals can only go so far. Information can always be copied. Try to keep your data within regions that have high ethical business standards and robust legal systems that pursue violations and enforce privacy laws.
This will mean paying a slightly higher fee. But, if security is a concern, it will be worth the cost. If location information is not advertised, call up the transcription service and ask where they intend to send your data. Starting prices less than £0.75 per recorded minute are a good indication of offshore and unsecured labour.
When it comes to automated transcription software, be cautious about how exactly your data is stored. Look to see if you can delete your files once they are uploaded. If you cannot, then your digital exposure to risk is significantly increased. You should also investigate privacy policies and make sure that ‘automated’ actually means no humans will look at your data.
Different digital file formats offer you options between compression and data loss. But, file formats don’t actually matter that much. If your recording has particularly low audio quality, preserving data will help human transcribers use editing software to clean up the audio, and may marginally improve results delivered by automated services. The cost is a larger file that will take longer to upload.
If you have a lot of recordings with relatively clean audio quality, MP3 or AAC files are fine. Both are compatible everywhere, they are compressed and easy to use. If concerned about audio quality, keeping your data uncompressed or using a ‘lossless’ compression file format might be a good idea. The two most common and widely interoperable of these are WAV and WMA. In general, however, different file types should not be a prohibitive issue.
Compression becomes more important for video files — the size of digital videos can get really big, really fast. MP4 (MPEG-4), FLV and AVI are all common and compressed video formats that are safe choices.
You need transcription services. Why you need transcription services is more complicated. Some people need transcriptions for record keeping, others need to digest and analyse large amounts of recorded information. Listening to audio is slow, and analysis can be difficult without the information on paper.
It is critical for market research firms to create and access detailed accounts of focus groups and in-depth interviews. Video and audio recording is great, but it is slow to sift through. Access to reliable human transcription services that pick up small details, contextually understand ‘meaning’ and accurately distinguish between multiple speakers is key to the industry. When looking for transcription services for market research, you should place a premium on accuracy, contextual understanding, data security and fast turnarounds.
The legal industry generates a lot of records through depositions, interviews and discovery. It is critical for both record keeping, in-court evidence and review to generate a large volume of high-quality and timestamped transcripts. Under certain circumstances, third-party legal transcriptions are a requirement. Confidentiality, accuracy and due-diligence are of paramount concern. It is important to find transcription partners that will guarantee proofreading and access to transcribers with legal experience.
Academics and students take a lot of notes. Sometimes making a recording is the best thing to do. For review purposes, academics can benefit from the summary and notes options available from human transcription services. If every detail is needed, transcripts provide an answer to that as well.
Meetings generate notes. Transcription services enable detailed record keeping and the creation of overview summaries to update senior management. Trained notetakers can even be hired to record meetings as they happen. That can help HR departments maintain a level of professionalism and efficiency.
The television and film industry need transcription services for both subtitles and closed captioning, along with time-coded post-production scripts. Quality, detail and always-available services allow post-production teams to assess scenes on the go, without having to watch every second. Quality services are key to subtitles and closed captioning. Captions online make content more accessible, increase engagement in boost SEO.
There are two categories of transcription services: automated speech to text transcription software and human transcription services. To determine what type of transcription service is right for you, you need to ask yourself six questions:
If you need to transcribe a complex recording of poor audio quality, you can basically rule out audio to text automated transcription software. Multiple voices, recordings made over the phone, background noise, complex subjects, accents and (particularly) poor audio quality will produce garbled and often indecipherable content when using automated transcription services. If only one or two of these issues are present, the transcript can be saved by going back through and editing it yourself. Often, however, the data is completely worthless.
If you need a professional and accurate transcript, you should expect to spend some time editing a transcript produced by automated speech to text software no matter how ‘ideal’ the audio file.
If you are having sensitive information transcribed, you should consider high-end human transcription services. Higher fees mean better-paid transcriptionists who are more likely to adhere to professional standards. They are also more likely to be based within your own country. This will improve your ability to pursue privacy violations, and decrease the likelihood that such violations occur.
Sometimes you don’t want a verbatim transcript. Sentences that are easy to understand in conversation can be perplexing to read. Often, part of the motivation for transcribing a recording is to increase the speed at which that information can be digested and analysed. If that is the case, you will benefit significantly from an edited version of the transcript. A simple summary of the recording might be even more useful. When using human transcription services, these options are all available to you. As you move away from a verbatim transcript, costs often decrease. If that is what you want, it is a win-win.
If you are working on a tight budget, you can sometimes find yourself in a jam. Ultimately, quality transcripts are difficult and time-consuming to produce. If circumstances allow you to make recordings in-person with a quality sound recorder, and you can get participants to speak clearly, slowly and one at a time, it may be possible to use cheap automated services. However, you should still be prepared to spend some time editing the transcript.
If transcribing every word isn’t critical, some human transcription services offer ‘notes’ style transcriptions or summaries at lower costs. Accepting slow turnaround times will often bring lower prices. Other transcription services offer bulk discounts or subscription options that will deliver higher value per transcript if the service is used regularly.
If your recording is not suited to automated speech to text transcription software and you have no budget to engage with human transcription services, you can always try to transcribe the audio yourself. Platforms such as oTranscribe provide a free interface that you can use to alter the speed of playback and pause/rewind without leaving the text document. As a guideline for what you are getting into, quality typists with experience transcribing generally spend about 4 minutes transcribing 1 minute of standard-quality audio. Be prepared and allot yourself the necessary time.
Now that you know everything there is to understand about transcription services, you need to pick a partner. If you have a high-quality/simple recording, are on a budget and have time to experiment, try the free versions of automated services to see if they work. When that fails, look at human transcription services.
If you plan on having a lot of audio transcribed or have started a new job where transcriptions are going to be a regular part of your life, put in the effort to make a good choice now. Look for opportunities to get bulk discounts or sign up for priority services. If you just have one thing to transcribe, just pick someone. If security and detail are not an issue, pick the cheapest service you can find. You just need to make sure that you aren’t going to get price gouged for an unexpected reason. If you care about specifics, context and data security, spend a bit more on the premium end of the market. No matter what, always do your research and contact services directly if you have questions.
If you’d like to send us an enquiry, or request a quotation, please fill in the form. We aim to respond to all email within 2 hours.
Alternatively please call 0207 928 1048 or use our live chat if you would prefer to speak to one of our team.