Posted by: patenttranslator | April 3, 2014

The Almost Incredibly Dysfunctional Modern Translation Industry


The translation industry, if we want to call it that, has undergone many changes in the last two or three decades. These changes may be invisible to young translators who may assume that the way things are now is the way things have always been.

But I remember that things were very different indeed when I was a newbie translator in 1987. The main difference between the translation industry then and now is that while back then translation industry was still very much about the art and craft of translation, now it mostly about the art and craft of buying and selling of translations – buying as low as possible, and selling as high as possible.

There were no huge translation agencies producing tens or hundreds of millions of dollars in revenue for their monolingual owners (who by virtue of not knowing any foreign language don’t understand the first thing about translating) every year through thousands of employees in dozens of offices.

I am looking at a list of Top 100 “Language Services Providers” (the newspeak for translation agencies, designed to hide the true nature of the business entities, which is that of a broker), compiled by a company called Common Sense Advisory, Inc.

The list says absolutely nothing about the countless invisible translators who translate and thus create millions of dollars in profit from which the monolingual translation agency owners and thousands of employees of the translation agencies can then live. This profit from the work of these unmentioned translators is also used to pay the rent for the offices, advertising costs, and all other associated costs.

But translators are not important in the modern model of the translation industry and if they are mentioned at all, it will be only in a single sentence on a website which says something like “We work with 3,000 (4,000 …. 5,000 or more) highly qualified translators”.

In the old days, before the advent of the mammoth translation agency model with hundreds or thousands of employees and many offices, small translation agencies, and it is almost comical how small they were compared to the monstrous structures that translators are dealing with today, in fact did try to work only with highly qualified translators because that was the key to survival in a marketplace where quality mattered.

Or so the translation agencies of yester years thought. That was why multilingual operators who were usually running the old type of translation agency were always on the lookout for the best translators out there, translators with degrees and experience in a given field, usually slightly weird people who loved their slightly weird occupation. Once they found such translators, they would work with them for many years, and they would pay them handsome rates for good work, because back then, the people who worked in these translation agencies were still able to tell the difference between good work, not-so-good work, and total garbage.

But all that changed with the advent of the mammoth translation agency model around 20 years ago. Most of the people who are now working in the large translation agency model, called project managers (or PMs for short) can’t really tell the difference between good work and not-so-good work because they almost never understand the languages from or into which something is translated. They can spot total garbage if the target language is in the one language that they do understand, but that’s about it.

The prevailing philosophy in the new translation agency model is that quality is somewhat overrated anyway, which is quite a logical conclusion considering that the people running things in this model, called PMs, can’t really tell a good translation from a bad one.

Quantity, on the other hand, is very important in this translation agency model because quantity can be measured and measured quantity, once there is a lot of it, equals higher and higher levels of profitability.

In addition to increasing the quantity of “production of translations”, another way to achieve higher profitability is to decrease the pay for the people who do the translating bit in the complex transactional process created by the modern and revolutionary translation agency model.

This relentless pressure on rates paid by agencies to translators has been further intensified by judicious use of modern technology by the translation agencies in the new translation agency model.

Internet, computer technology, globalization and the never-ending economic recession (we are not supposed to call it a crisis) are the four prongs of a giant fork used in the modern translation industry model to push the rates paid for translation to those invisible translators who work quietly in the background as low as possible, and then to keep them as low as possible.

This is similar to what has been happening for quite some time to many other white collar professions that used to employ highly paid professionals in a number of fields. As per this article in New York Times from 2003, complicated X-ray images that used to be analyzed by American radiologists who were making up to 350 thousand dollars in United States are now often analyzed by cheap Indian radiologists for 25 thousand dollars a year.

It’s good money for the people in India, and the private health insurance companies and hospitals in United States get to keep the difference in pay because the medical costs in United States have gone through the roof since 2003.

Translation rates are also squeezed by CAT (computer-assisted translation) software which is used by translation agencies to greatly reduce or simply refuse to pay for words repeated in a translation (a concept called “full matches” and “fuzzy matches”). The difference in the cost is again mostly kept by the translation agencies.

A new concept pushed by a brand new model of translation agencies is human-assisted machine translation, which is something that could use a new abbreviation.

HAMT? Or can somebody think of a better and snappier abbreviation?

In the old days, humans used to be assisted by machines. These days, machines are assisted by humans.

Innovative new startups, and new ones are being created as I am writing these words on my silly blog, are based on the concept of crowdsourced editing of machine translation, namely the idea that the minor imperfections of machine translation, such as when they make no sense whatsoever, can be corrected by humans who don’t really need to be translators at all as long as they have some knowledge of two languages, and who should be able to make these corrections on their cell phones for something like 1 or 2 cents per word.

According to a youthful owner of one such startup called “Flitto”, [Machine translation] … “has been unable to shake inaccuracy issues. Nobody uses Google for translating business documents. We believe we can outdo Google in this field. Flitto is a great way to make money with your free time. While you’re sitting in the toilet, you can utilize the time efficiently by translating and earning points”.

Maybe it’s just me, but it seems to me that the modern translation industry is now completely dysfunctional.

It seems to me that people who perhaps know another language to some extent, who have no credentials and no experience in the translating field, are likely to just further mess up the garbage that was originally produced by machine translation instead of fixing the “tiny problems”, for instance when they are working on a mobile phone while sitting on a toilet.

Maybe it’s just me, but I think that the translation industry, with all of its modern tools, from shifting of work away from educated and experienced translators to people in third world countries who may be translating from languages that they don’t understand very well into a language that they have not really mastered, to machine translation and working on a cell phone while sitting on a toilet, is now a complete joke.

I think that the inevitable result of these trends is that the modern translation industry model will be producing mostly just heaps of garbage.

I also think that the best thing that customers who need real translations for really important projects should do would be to stay away from the modern model of the translation industry altogether and work instead only with small translation agencies who still believe in the old model, or directly with translators.

Otherwise, there is a good chance that the translation that they just paid good money for was first processed by machine translation software, and then edited by a dude who says that he knows some German, or Japanese, or Korean, and who was editing the machine-translated detritus on his cell phone while sitting on a toilet.




  1. “…the translation that you just paid good money for was first processed by machine translation software, and then edited by a dude… while sitting on a toilet.”

    A nice way to make some money when one is retired, isn’t it?

    There is quite a difference between being classy and being reliable. Machines are reliable, but never classy. You need to be able to refrain yourself from sic-king on someone’s non-native language performance in order to be classy. But the dude sitting on a toilet can be highly reliable editing some machine spews – and with native performance, of course.


  2. @Wenjer

    Even if he has a diarrhea?


    • It could happen to anyone.


    • Oh, BTW, I like both video clips you chose this time.


  3. I do agree with all of this, but it occurred to me that there is one important factor you are not taking into account: the fact that the quantity of translation has changed. Before the internet, translation was a professional affair. Legal translation is a good example. You needed to have special qualifications, well, you know the rest. Nowadays the pm says ‘it’s legal’ if they can be bothered to focus on it and you are expected to be able to do legal translation as a matter of course – although by the standards of the real legal profession, this is almost impossible. As a Japanese translator, you probably know that during the Tokyo trials after the Second World War, translating Japanese law into ‘international’ (American) law and vice versa was an insurmountable issue; but today, what the hell, as long as you get the product out there then what the hell. If you yourself fail, there are tens of thousands of other people who can take the fall next time. The point is that there is much more translation being done, and that is the central reason behind all these issues of quality, the appalling business models, and so on. Not only that: the demand for translation is steadily increasing. As internationalists, we should be in favour than that I suppose.


  4. “Oh, BTW, I like both video clips you chose this time.”

    I know.

    You are so predictable.


    • In other word, reliable.


  5. @Christopher

    I agree, the need for quantity has made quality much less important.

    But still, est modus in rebus et sunt certi denique fines.

    There is legal translation and legal translation.

    I translate Divorce Judgments from Japanese and Divorce Decisions from German relatively frequently, but I would not be able to translate new Japanese or German law, even after all these years.


  6. Well, Steve, I told you the MT post editing world would be here soon two years ago. Now you are writing: “I think that the inevitable results of these trends is that the modern translation industry model will be producing mostly just heaps of garbage.”


  7. @Nate

    And two years ago I told you that it would not work, didn’t I?


  8. I am having difficulty imagining the dude on the toilet thrashing out 3,000 perfectly typed words per day on his mobile.
    I start thinking of things like,
    “Oops, not ‘now’, but ‘not'” added in a text message as an afterthought referring to a (fictional) sentence which was intended to read “Toilet translations are not permitted.”
    Not in my office, that’s is for sure. 🙂


    • I swear I did not type that on a mobile. I meant “that is for sure”, of course. Imagine what I would be like on a mobile! :p


  9. @Allison

    I have to proofread everything I write several times to catch all or most of the typos.

    With the posts on this blog, I usually publish them first, and then proofread and try to catch the typos once a post has been published. I find it very hard to see everything that I need to see in that little editing window on WordPress before a post is published.

    I proofread translations the next day whenever I can do that to catch all mistakes and typos.

    If I were to work on my cell phone while sitting on the toilet, the result would be …. Sheisse, Merde, Kuso!

    But this is precisely what the inventors of the new, revolutionary translation techniques linked in my silly post are suggesting.


  10. I really enjoyed reading this post, Steve. You hit your stride.

    Thankfully, there are still some good agencies out there. I hope they don’t get driven out of business.


    • “Thankfully, there are still some good agencies out there. I hope they don’t get driven out of business.”

      It may be that two types of translation agencies will emerge in the end from the translation industry morass: the old type, the type that values translators and their work, and the new type, the type that is based on endless exploitation of nameless warm bodies working on a cell phone in a toilet somewhere.

      If real translators stick with traditional agencies, the new type will have to work mostly with zombie translators, which in the long run is likely to affect their bottom line in ways that might not have been anticipated.


      • I tend to agree with your prognosis.

        (As an aside, it is, of course, economically more profitable to work while sitting on the toilet: that way you don’t have to take potty breaks. I wonder why some enterprising company hasn’t come up with a “toilet desk”…)


  11. Please have a look at this job offer and the word-count system: 2500 new words – 13.000 repetitions? I am speechless.

    Dear Translators & Proof readers,
    We have a new project that involves the translation of software in Passolo for a client that deals with medical equipment.
    The source text is a manual which serves as a reference guide for trained personnel (e.g. audiologists) conducting the programming of audio processors and associated testing of implant users.
    We require English into French translators who use Passolo that are available to start today or at the latest tomorrow, Friday 4 April 2014.
    We also require proof readers who use Passolo for 5 languages (English source):
    Chinese Simplified
    Portuguese (Brazil)
    Proof reading will start next week.
    Word count: approximately 2,500 new words and 13,000 repetitions.
    If you are interested and have the relevant skills and experience, please e-mail
    – Your CV with details of your professional experiences (please specify your mother tongue)
    – Your translation rate per new source word
    – Your hourly rate for proof reading

    Please include the identifier “Passolo project – translation and proof reading (English into 5 languages) – Kasia W” in the subject line.
    I am looking forward to receiving your e-mail.
    Please note that this e-mail is an availability request only at this stage, and is not a request to start work. The job has not been assigned until we send a job brief via e-mail or i plus, or a written confirmation to proceed.
    Thank you!
    Best regards,
    Kasia Woźniak

    Software richiesto: PASSOLO.

    Scadenza: 10/04/2014 23:00


  12. @ Vincenza

    Kasia may have problems with the rate she wants to pay even with translators in Poland, which is where she is from.

    Another great thought that occurred to me, Kasia (Kasha) means “porridge” in Russian, and there is a Russian idiom that says “Kasha mat nasha” (“Porridge is our mother”) which is what Russian people tell to themselves when tasteless porridge is again on the breakfast menu.

    But this Polish Kasia is definitely not our mother.

    Liked by 1 person

    • On considering the expertise and language pairs required as well as the use of a specific software as being the prerequisite condition, they seem to have no chance in the Chindia area as well.
      Viel Spaß to us all & Ciao!


  13. “As an aside, it is, of course, economically more profitable to work while sitting on the toilet”

    Google just might try it if somebody there reads my posts.

    Since they already have free coke and free meals for their employees, Google workers tend to have more potty breaks than employees of other companies.


  14. @Rob

    The desk is perfect for Flitto translation technology.

    Now I understand what the saying “everything is in the toilet” means.


    • Ha ha 🙂


  15. […] “The translation industry, if we want to call it that, has undergone many changes in the last two or three decades. These changes may be invisible to young translators who may assume that the way things are now is the way …”  […]


  16. HAMT (Human assisted machine translation)…….

    I vote for either “Hum-ass” or “Has-Mat”


  17. I too started to translate back in 1987, mainly for publishing houses and using an electric typewriter. They even provided you with the sheets of paper you should use. So you had to really think the sentences before you typed them down!
    The job has certainly changed. I recently restarted as a freelancer after 13 years holding an in-house position (as translator, proofreader and QM) in a small translation company. Now I admit I have among my clients one of the biggest global translation companies, with offices in Hong Kong, Honolulu, New York, Barcelona… I use Wordfast and I accept their wordcount analysis.
    You certainly have to somehow depersonalise yourself because, once you send a translation you cannot know what path it will follow, or if it will end up in the hands (or under the eyes) of a recent graduate who will destroy all your effort for consistency, terminology accuracy and style. On the opposite side, when the job I perform involves a proofreading, I always try to be extremely respectful with the translator… which sometimes is really hard job because I read translations I would never have imagined.
    Is it the sign of the times? Maybe… but, I can see an anology here: when CDs started to replace vinyl records, I told everybody who would listen that I felt their quality of sound was lower and poorer than that of vinyls. Vinyl records almost disappeared to give rise to an increasingly omnipresent digital music. Now vinyls are slowly coming back for those willing to listen to something more than binary-code music.
    Perhaps we can be optimistic and think that, at least for some people, quality after all does and will matter.


  18. @Medusaurelia

    The problem with quality is that it matters only to people who can tell what quality is.

    In a world inundated mostly with crap everywhere one looks, this would be fewer and fewer people, and I am not talking only about translation here.

    But I agree with you that just like there is a market for shops specializing in vinyl records now that not that many people still buy CDs, there is also a market for specialized translation services now that translation agencies have been inundating the market for years with CAT-processed and MT-processed and human-assisted crap.


  19. Kevin Lossner in his article In HAMPsTr We Trust? coined the term HAMPsTr (Human-Assisted Machine pseudo-Translation) to describe just that, and I find it to be very fitting, especially when imagining the millions of hamster-like nanolators running in their virtual toilet shaped wheels while going nowhere.

    In my opinion there isn’t such thing as a translation industry, and if there is, the term refers to the lower (some call them bulk) market segments which are filled with all of this nonsense.

    The demand for translation increases, but it has increased and will probably continue to increase in the lower segments more than in the professional ones. Some clients will graduate from the lower segments and advance to the professional ones, others will remain in that toxic cesspool. This divide is not unique to the translation market, but the professional segments (which are comprised from professional translators and professional practices, i.e. the older type of “agency”) should work harder to separate themselves from the “industry”.

    I think that too much energy is wasted on trying to convince people that translation matters, that you can’t get quality stuff by turning to those charlatans, etc. It is wasted because the average client is a layman that doesn’t know anything about translation (and sadly, so do many who pretend to represent the profession) so the reasoning, as solid and robust as it may be, is wasted on them. Some will learn the hard way, but when they do, they need to know where to turn to get a professional service – and I think that in this respect the professional market-tier is still lacking.

    The new agency business model is all about brokerage, and its structure creates fixed costs: the more they expand, the larger the fixed costs are. To that end, they need to drive volume to cover those expenses, the more volume they generate, the better it is for them. Although it is the core service that they offer, in their book the translation “cost” is just another fixed cost they need to keep down. Quality is not even a concern for them, because this type of business is always about the bottom line. This is why so many of them are turning to MT and Crowdsourcing (oh, sorry, they are starting to call it Social Translation).

    Professional practices and independent translators, on the other hand, have a completely different business model. Their fixed overhead costs are of less importance, and their main asset is their intellectual capital, which is the core element that allows them to do the work that they do. For them it is not about dealing in fixed costs, it is about time spent on a project and getting and charging a reflective and appropriate fee for the effort involved in that project.

    This is a crucial, arguably unbridgeable, difference between the two business models, and why the modern type of agencies and translators are not partners that work towards a common goal in a symbiotic-like relationships, but more of a predator-prey type of relationship. This is also why true professionals should limit their exposure to those organizations to the absolute minimum and instead work on increasing their exposure to the content creators who just might actually appreciate and value their intellectual property, expertise, skill and advice.

    The problem is that those brokers and some of the technology developers (sometimes they have a direct business affiliations) that I like to call the technology-lobby have embarked on a vast social engineering campaign that aims to change the perception of quality for the purpose of increasing and expanding their market segments and profit margins (for the above reasons), but this is a topic for another discussion, I guess.


  20. “The problem is that those brokers and some of the technology developers (sometimes they have a direct business affiliations) that I like to call the technology-lobby have embarked on a vast social engineering campaign that aims to change the perception of quality for the purpose of increasing and expanding their market segments and profit margins (for the above reasons), but this is a topic for another discussion, I guess.”

    And I think that they have been quite successful, or at least moderately successful, in making people believe, including many translators and at least some customers, that instead of emphasizing specialized knowledge gained from decades of education and experience, the answer to the problems represented by the giant mountains of crap that have been generated by this translation industry in lieu of real translations is in the same technology as the technology of a slightly earlier vintage that created the mountains of crap in the first place.

    It is a seductive concept because we are all led to believe by the worldwide religion of Profit Ueber Alles, that every problem, including the survival of the human species, can be solved with judicious application of the right kind of technology, once the technology has been tested and optimized for a given market segment.

    Taking a critical look at this insanely greedy (rather than insanely great) structure and replacing the parts of the structure that are responsible for most of the crap is the last thing that these brokers who are the modern translation industry, is the last thing that they would be willing even to consider because it would make apparent even to people who know very little about translation, i.e. our customers, that the emperor has no clothes.


    • I think that they have succeeded on two fronts:
      1) In convincing both end-clients and translators that they are dependent on them;
      2) In changing the perception of quality, making people (again end-clients and translators) believe that they can go only as so far as the technological limitations enable them to, as if there aren’t decades-upon-decades of great translation history.

      The latter is actually something I’ve start to outline for an article on my own very little blog, but if the offer to write a guest post still stands, I think that your blog is a better place for it.


  21. “The latter is actually something I’ve start to outline for an article on my own very little blog, but if the offer to write a guest post still stands, I think that your blog is a better place for it.”

    I didn’t know you had a blog, but I my offer still stands, and my blog would be as good a place for such an article as yours.


  22. I agree with much of the article and Shai’s analysis; I consider translation to be a profession. There is a translation “industry” but not one that professional translators are a part of, pretty much as doctors are medical professionals but don’t work in the medical “industry”. I also believe it’s pretty much a case of horses for courses: there will always be a demand for both high-end specialised translators who know what they’re doing with or without MT, and bottom-end translators for the “translation industry”. While there is pressure to drive down rates for high-enders, there is little competition in terms of quality and others with similar skills.
    Another problem with the pervasive technology lobby (and the CSA set) is that they are now endorsed by bodies that are representative of translators; journals I am a subscriber to as a member of those bodies focus almost entirely on the “industry” side of matters and how to increase productivity as opposed to “how” to translate, and what the issues involved in translation are. It’s not about reading academic theses because, as interesting as Translation Studies is, it too is fairly divorced from translation practice, however there seems to be an official, perhaps sponsored, line that technology and translation solely “as product”, as opposed to “as process”, matter. I’m all for progress, but I prefer quality all the way. I believe that is professional and what only a professional can offer, as mentioned above.


  23. Good stuff Steve! I’d say more but since I haven’t yet been replaced by any of the entities that you refer to I’ve got to get back to work PDQ 🙂


  24. @Aisha and Alchemie

    It is interesting that no fans of the translation industry among translators or representatives of the translation industry weighed in so far to defend the revolutionary technologies which according to the industry will make it possible to translate huge amounts of data without compromising quality and at a reasonable price.

    If that is their argument, and if the argument holds water, shouldn’t they be able to defend it in an adversarial environment?


  25. My post became a catalyst triggering 2 interesting discussions on proz:
    (Thanks, Jeff!)


  26. Great post! I am a young translator but I absolutely agree with you. I have already left a couple of agencies because of money matters: one of them shifted their bank commission on me (extra 22 EUR per transaction), another started bargaining about an impossible discount after 5 years of cooperation. The point is that I am not frustrated much about our break, but I am not sure that my place will be taken by a careful translator and both these agencies do not have a sustainable system of quality control, everything is on a translator’s head. Sometimes I outsource projects and I see how many so-called translators are non-professional, so the chance for these agencies to find a good specialist is not very big. And then we have awful literary or even machine-like translations which are impossible to read or even understand. It’s not a big deal when it is a toy description, but when it comes to serious manuals…


  27. Well you are indeed correct, the industry has changed since 1987…with the internet you can now deliver digital content in the blink of an eye, market to a global audience, manage a global team and have access to almost every written document on earth. And what was once an “art and craft” is now a maturing industry, that needs sustainable business models. Yes, the free market will drive prices down but only for those that provide mediocre quality, but it will also create a higher demand (with higher prices for translators) who are experts in their industries and who have a real delivery strategy. For companies to compete globally they will need to effectively communicate globally. A crucial part of that is to have an effective language translation strategy that includes partnering with language service providers who not only are experts in the translation world but thoroughly understand their clients industries. Those that choose to use 3rd world providers, or will go to the cheapest bidder will quickly find themselves unable to compete, and eventually out of business. So, in closing, times may have changed, but definitely for the better, at least for the translation buyer.


  28. “A crucial part of that is to have an effective language translation strategy that includes partnering with language service providers who not only are experts in the translation world but thoroughly understand their clients industries”

    So what do you understand about my client’s industries that I don’t understand, I wonder.


  29. An excellent and very honest State of the Profession Address, dear colleague! I fact, I’m often missing this kind of honesty in discussions among fellow translators. An overwhelming majority prefers to look the other way, hoping that “the market” will sort this out, all on its own (an attitude which IMHO is part of the problem rather than of the solution.)

    But leaving the purely money-related pressures aside for one moment, I do agree that part of the problem is due to changed perceptions — engineered or not — and in fact a change in the overall attitudes towards many human endeavours in wider society. The ideology of efficiency now has an almost total (not to say totalitarian in some parts) prevalence over everything else. Things used to be different, more balanced, even 20 years ago, at least in Europe. So what you’re describing is part of a larger problem IMHO.

    I’m also finding that the prevailing concept of language is flatly wrong, or outdated, sadly even among many language professionals. E.g. the tacit idea of language being some kind of “input stuff” or mere “information”, even among many translators, has huge consequences for the acceptance and use of CAT, MT and other technologies. Portraying language as some kind of fixed (though complex) structure has the advantage that it fits the MT cycle, it plays into the hands of the “industry”. But it’s still wrong, and so are the results — pure alchemy, quite literally.

    I agree that a large part of this “industry” has become dysfunctional. I would say that the translation process does not scale too well, or not at all in many regards, and it shows. Many things have spiralled out of control, layers of complexity have been heaped upon us which are counterproductive (e.g. certain standards), plus they undermine our confidence and financial standing. I don’t know whether we can rebuild trust and honesty, and more direct and functional relationships, but we must give it a try. Somehow it must be possible to reinject some reason into the whole thing if it is not to implode. I’m not hoping for the latter actually, I’m not awaiting the system crash “so that it will separate the good from the bad”, but as we have seen, it is certainly a possibility. We can’t go back to the eighties but certainly past experiences can serve as an orientation to see what has been lost over the years, so I’m glad one of the older colleagues is speaking up here.

    Thanks again for your clear words.


  30. […] The translation industry, if we want to call it that, has undergone many changes in the last two or three decades. These changes may be invisible to young translators who may assume that the way things are now is the way things have always been.  […]


  31. […] this changed with the advent of the greedy corporate translation agency model in which translators are considered easily replaceable indentured servants who have no rights rather […]


  32. […] which would you learn, and why? 11 Product Names That Mean Unfortunate Things in Other Languages The Almost Incredibly Dysfunctional Modern Translation Industry Creating and managing your own contact list of reliable colleagues What You Need to Know About […]


  33. […] platform” on the Internet and translate from their home on a laptop or even on a smart phone, for example while sitting in the bathroom. What an ingenious multitasking concept? It must have been invented by a compulsive iPhone user […]


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s


%d bloggers like this: