This is post #4 in my ongoing series of questions from the webinar on “Getting Started as a Freelance Translator” that I presented for the American Translators Association in December.
A participant asks: For a translator who doesn’t use CAT tools, what is the average turnaround/output in words per day?
Short answer: If you want to work for agencies, you probably need to translate 2,000-3,000 words per day in order to meet their deadlines. If you work for direct clients, you can work more at your own pace. I know translators who average anywhere between 200 and 1,000 finished words per hour, so obviously your mileage will vary!
Longer answer: First, I’m not completely convinced that CAT tools save translators a lot of time unless the project is extremely repetitive (i.e. updating a previously translated document). When I use WordFast or OmegaT, I find that my translation speed increases by about 10% because I’m not constantly finding my place in the source document and glancing back and forth between screens, but I also find that my editing time increases because my writing is more “chunked” and does not flow as well. I avoid using CAT tools on anything that is for publication, but that’s another post!
If I had to put a number on it, I would say that the average translator produces between 400 and 600 finished words per hour. However, most people can’t translate for 8 hours without stopping, and most of the time you’ll encounter a section in the document that you have to research or read about, so you’ll slow down. In addition, most translators’ work speed varies enormously depending on the subject matter and format of the document. Years ago I used to translate market research surveys that were so repetitive I could listen to audio books while I worked and still produce about 800 finished words per hour. However, this also meant that the work was not very stimulating/fulfilling/intellectually demanding and therefore not very satisfying. I’ve worked on really complex legal documents that involved multiple cross-references and layers of meaning, and I’ve dipped as low as 250 words per hour.
As I mentioned in the short answer to this question, I think that agencies expect that translators will produce 2,000 to 3,000 finished words per day, and in general you have to be able to translate that fast in order to make a healthy income on agency rates (of course there are exceptions!). I’m not in the “faster is better” camp, but if you want to translate faster, you could:
- Work on your typing. It sounds simplistic, but a lot of translators could translate faster if they typed faster. Recently, someone doing a research project on translators’ typing speeds asked me to take this typing speed test. I scored 83 words a minute (eternal thanks to my 10th grade keyboarding teacher!); according to the typing website, this means that I save 5.5 hours per 10 hours of typing as compared with “the average typist” who does 36 words per minute. Moral: typing speed matters!
- Use speech recognition software or hire a transcriptionist. Most speech recognition software and most transcriptionists should be able to handle 80-120 words per minute. If you’re at that 36 word per minute average, this could really increase your productivity. I haven’t tried speech recognition software because I just don’t think that fast! And I don’t mind typing. But most of the ultra-productive translators I know (800-1,000 words per hour) do use speech recognition software.
- Concentrate on your specializations. I think that the best way to increase your translation speed is to become really knowledgeable about your areas of specialization and stick to them. Once you start to know the terminology, sentence structure, typical phrasing, and even your regular clients’ writing styles, you’ll really speed up.
Any other thoughts on translation speed?
Ahhh…interesting question!
Personally, I can hit “burst” speeds of 1,000 words an hour and regularly do urgent newspaper articles for corporate press reviews at about that rate. However, I could not and would not want to do it all day long or every day of the week. A couple of “bursts” a week is enough and, given the high rush rates, is a nice income booster.
More generally, I tend to produce around 2,000 words a day. Sometimes more (on docs like RFP submissions or marketing briefs that are heavy on the “blablas” but light on actual meaningful messages), sometimes less (on more creative work like ads and brochures where every word carries a lot of meaning and must be thought through and revised thoroughly).
At a recent conference at the University of Grenoble, one thing that emerged as a real stumbling block to translator productivity was a lack of mastery of the Windows environment and basic MS Office software. The ability to quickly and efficiently navigate between windows, perform web and desktop searches, and manipulate different file versions seems to me a good way to achieve higher speeds and reduce fatigue. I know I need to get better at this.
I am with you 100% on the typing (we must have been in high school at about the same time in the US…god bless those old schoolmarmy typing teachers!). Saves time and reduces fatigue.
I cheerfully exclude myself from the group of speedy translators. I am actually quite slow, and since I don’t work for agencies, I have very manageable deadlines, allowing me to work on several projects at a time. If I had to put a number on it, I’d probably say 1,500/day, more if needed.
Yes, typing speed is key. I have also scored in the 80s, and I used to have a boss who typed 150 (he won the national online typing contest). We used to stop by his office just to watch him type – quite a sight. I didn’t have typing in high school, and for some reason I type with 7 fingers, but I am quite speedy.
One more thing I can think of: minimize distractions, such as checking e-mail every five minutes (I am horrible at it, and do it, too).
I think subject mastery, format and the system familiarity issues Sara mentions are the main factors. The belief that typing speed is a primary bottleneck is a superstition that persists, perhaps because of the eternal confusion between translators and bilingual secretaries and the overlap one often finds.
Just for laughs, someone with a desire to do a bit of research could pack a room with translators for a given language pair, test their typing speed and then have them translate a few texts of varying difficulty. If there is any correlation at all with typing speed, it might be found with the simplest texts, but even there I have my doubts.
Let’s do a bit of simple math. Say you type 50 wpm. How many words is that in an hour? 300 you say? Yes, I’m sure you’re a fine translator. Check your work. 3000? That’s better. Now when was the last time you heard of someone translating 3000 words per hour?
The bursts that Sara cites correspond to 1000/60 = 16.7 or a bit less than 17 words per minute. A blinding speed of 17 words per minute while translating! Well, she’s doing better than me; most days I barely manage 10 wpm, but I get by.
So what takes all that extra time? Mostly thinking and research I presume. When I stop to consider the best way to phrase something, my typing speed does not apply. I can indeed think while I type, but I’m usually thinking about the next sentence while I type the previous one. Or I re-think a sentence as I type it: a real speed killer. Yet my 400 to 500 wph average gets me through the day faster enough.
Corinne’s last tip about knowing your subject is absolutely on target. Give me a chemical procedure or safety information and you’ll probably see that Holy Grail 1000+ wph even with my arthritic fingers. Yet my typing still plods along at about 45 wpm.
We aren’t secretaries thoughtlessly retyping handwritten letter from the boss. Let’s stop measuring ourselves by their standards.
Although I’m only a relatively quick typist (65 wpm), I don’t mind that because it allows my to multitask and check my translation at the same time. I use CAT tools for almost all my translation work, so the per sentence structure works for me, I proofread my translation again before closing the segment and that really saves time at the end when I have to proofread the whole text. In general, I’d say that I translate 500-600 words per hour, 2500-3000 words per day, for extremely tight deadlines, that can go up to 4000, but for a few days only. Replying to emails, issuing invoices etc. take up a lot of time during the day and I’d have to lose sleep if I ever wanted to translate more than my usual. Plus, I don’t think it’s a good idea to translate after a certain amount of work time, because you can’t be as concentrated and you’re prone to mistakes which your tired eyes won’t catch (at least mine don’t :-)).
Thanks for the nice post Corinne!
Productivity analysis is one of my favourite areas too. Minimising the need to break off in order to look up a word (which comes with subject mastery and experience) is quite important, as is minimising key strokes through familiarity with one’s tools and working environment.
Something that I find to be fundamental, though, is “managing” one’s mind, both by keeping fresh (proper sleep, regular breaks, sufficient physical exercise) and by managing attention – reducing distractions, keeping Twitter in its box (or nest, probably that should be), and trying to avoid switching back and forth between multiple tasks, which I find frazzles my brain completely if I’m not careful. Meditation can be very useful to that end; once I discipline myself to do this on a regular basis, I’ll be well away :).
In my first in-house job in Italy, the basic rate of pay was awarded for a minimum of four “cartelle”/hour (c. 900 words). Fortunately the material was generally simple “news and views” pieces, but the sharp pay differential if you dipped below that definitely kept you motivated to keep it up, even over a four hour shift.
I’ve remained quite quick on simple tasks, but like most people I level out at around 2500 words per day.
Recently I began experimenting with Windows speech recognition software, which isn’t all that bad, actually. Eventually I’m aiming to dictate all my translations in a bid to avoid arthritis and back pain….and it just seems much more relaxing than bashing away at the keyboard all day!
Thanks, Corinne. I wholeheartedly agree with you on CAT tools – helpful with repetitive texts for internal use, not all suitable with materials intended for publication!
Great advice on productivity analysis. I have a feeling that many translators overlook the importance of typing speed. I’ve always been reasonably fast at typing in English (65wpm on the test you linked), but only since I began doing English-Russian translations have I really appreciated its importance. Non-Latin keyboards take some time! I’m at 38wpm in Russian, which is acceptable for the volume of work I have. Still, it leaves a lot of room for progress!
I’m very intrigued by the idea of using speech recognition software, although I can’t help but wonder whether time spent on editing the final product would increase significantly. I’m a very visual person, and what sounds like a good sentence when spoken may turn out quite dreadful on the page. But of course it would be wonderful to avoid the backaches, muscle stiffness, carpal tunnel syndrome and any of the other ailments associated with sedentary work.
I can’t argue with Kevin’s careful analysis of typing/thinking/translating speeds. But to me the greatest benefit of knowing how to touch type is less fatigue and thus perhaps overall slightly faster speeds over the course of a day…Not to mention feeling better at the end of the day than if I had hunted and pecked for eight hours.
Sara, what you mention here is ergonomics, and that is very important. But it has nothing to do with typing speed. It’s entirely plausible that as a relaxed touch typist you will be less fatigued and think more clearly at the end of the day. Purely theoretical for me, as I hunt and peck unfortunately. But even long before I started to translate commercially, I noticed that my reports usually got written faster than my colleagues who were fast typists. The limiting factor was more how fast our minds processed the data we described.
My life would definitely be better as a touch typist. I would probably spend most of my time watching the magpies quarrel in the courtyard outside my window as I type, but whether I did so at 40 wpm, 60, or even 100 wpm would make little difference at the end of the day. Superior posture, however, very well might.
Speaking of which, I think a lot about that standing workstation (or treadmill I think it was) that Corinne once described. I think that may be a change worth making.
The speed depends in my case mostly on the subject and on the length of the document. I translate between about 2 thousand to slightly over 5 thousand words a day. (God bless Yoshiko-san at San Francisco Convention & Visitors Bureau who insisted 28 years ago that I will never amount to anything unless I learn touch typing first! She was right)!
I don’t use any CATs and probably never will. Today, for example, I finished two short German patents, one about 800 words and one about 1,700 words, both were about forklift trucks. Short patents like this with lots of specialized terminology, in this case mechanical engineering, are very labor intensive for me because I keep comparing my English terms to what I can find online, (usually the WIPO site for German), and I keep changing the terminology until I am happy with it. Tomorrow I have a German patent about a data processing method, about 5,000 words, and I will probably finish all of it or most of it in one day because data processing terminology is much simpler then mechanical engineering.
One law firm has been sending me for several years Japanese patents about medical information processing technology, each patent between 10 to 20 thousand words. On things like that I can fly. There are only a few new terms that I need to confirm for each new patent and I can then go over 5,000 words a day. I prefer Japanese to German and I am faster when I translate Japanese compared to other languages. It’s partly because I know the technical terms better in Japanese than in other languages, but I think it’s partly also because the graphic form of characters can be processed visually in your brain faster than for instance a long German compound word, but it’s just a theory that I have.
I am also highly motivated by greed – if I get my rush rate, which is 40% higher, I don’t mind pushing myself to the limit, which is about the 5,000 mark. If all I can get is my regular rate, I make a lot of breaks, watch me a little teevee, stuff my face downstairs in the kitchen, read and write blogs, go to the gym ….
Which reminds me, I really should go to the gym today. I had two long rush Japanese patents this week and I was not exercising at all.
Mens sana in corpore sano!
To me the editing is much more time consuming than the translating. I have a 3-step process and step 2 (comparing the German to the English, line by line), is by far the most time consuming, and I can do about 3 pages of that an hour.
All said, I think that 2,000 words per day is a good limit. I certainly don’t want to compete with translators who race through and boast about their words per hour. I want to compete on quality, not volume.
Regarding speech recognition, I found Mac’s dictation program fast, but I had to do extra editing (lots of “hidden” errors in the text I ended up with).
“I have a 3-step process and step 2 (comparing the German to the English, line by line), is by far the most time consuming, and I can do about 3 pages of that an hour.”
What are the other 2 steps?
Just wondering if it is similar to what I do.
Steve Vitek
I am currently investigating another way of saving time: speed reading. As translators, we spend hours reading every day: source and target texts, supporting documentation, and obviously colleagues’ blogs! I am pretty sure educating our eyes so that we are able to scan through a text quickly may change the way we work. However I am not sure it is a safe practice when it comes to reviewing a translation or any other task requiring great attention to details… Still testing that!
I cannot translate over 1500 words per day. I think pretty slowly and do first drafts on paper in longhand. I then type out my work, print it out, and edit more with pen in hand.
Your response to a tricky question, Corinne, is one of the best and most detailed I’ve read so far. Thank you! This issue keeps coming up every now and then. I can only recommend that people translate and edit in there own pace, research as much as they need and – last but not least – have breaks. So far, that has proved to be the best way for most of my colleagues and me who’ve worked in this business for many years. As much as I understand that agencies need some kind of reference point, I’ve learnt from experience that it’s contraproductive to let them determine on the amount of words/pages a translator can manage in a day.
Thanks everyone for your really helpful comments; I think that they add a great deal to the original post since I can only give my point of view! Keep them coming and thank you for contributing!
The other two steps are: 1. translate, and 3. Proof read. The second step usually takes me as long as the first. The third step is the easiest and fastest.
I agree that 1,500 words a day max. is the right number for “complex”, or tough subjects (i.e. stuff you don’t specialize in). 2,000 to 2,500 would be my limit for stuff I am familiar with.
When I worked in the translation department of a business software company, we did gather statistics about translation speed. The projects were similar, and the work organized in a similar way (translations done with a CAT tool, then edited by a different translator before going to DTP).
We found that translation speed varied considerably, even within the same language team: the fastest translators did about 4,500 words per day, the slowest ones about 1,500, wiht an average of about 2,000 words/day.
For editing (bilingual editing), we averaged 6,000 to 7,500 words per day.
Now that I work again for myself (or rather, for our small translation company), I find I usually work faster: if I really need to I can get to 6,000 to 7,000 words/day (using a CAT tool, of course) of translation, and at least twice as much bilingual editing.
BTW, that 6,000 to 7,000 words/day of translation is a maximum speed. My average for translation, self-edited and ready to go to my editor is more like 3,000 to 3,500 words/day.
Hi. Great thread. I come down on the side of typing speed being a lot less important than research time and thinking speed. I say this because my output is pretty high (about 3000 words a day ongoing, 4000 wpd once or twice a month) but my typing speed is only about 45 wpm.
My research times are probably pretty low because I’m highly specialized and don’t have to spend a lot of time in the dictionaries or nailing down terms. I expect that balances my not so stellar typing speed. I’ll think about whether it might be worth doing some training though, after reading this.
Knowing windows shortcuts so that less time is lost to mouse movement is an underrated time- and wrist-saver.
Hello everybody!
Though some time has gone by since the last post on this thread, I’d like to add a tip: get a bigger screen! I have two 25″ screens, one for writing and one for looking for info (dictionaries, browsers etc.). Still, they’re never big enough 🙂 I’m planning to get a new TV for my living room and get the old one – a 37″ good-quality TV – to my desk. That way, I hope I can keep more windows opened at the same time and reduce the clicking/mouse-moving time. Furthermore, the bigger the screen, the bigger the font-size, with positive side-effects on one’s eyes.
Any similar technical tips?
Ciao!
For me, the crucial factors affecting output (using CAT tools) are:
checking terminology for an unfamiliar field (research in general);
typing speed in a familiar subject (if only I could touch type and have the time to practice – I am old enough to know that my handwriting speed is faster than my typing speed);
proof reading post facto (I usually prefer a night’s sleep before doing this so that i can have an uncluttered mind) ;
Size of screen or the use of 2 or more screens (the change from laptop to desk top or vice versa is significant).
I fully agree with Gianfranco regarding the screen size. I use 2 22″ screens one for the source text (especially useful for pdf files) and one for the translation environment with several Italian dictionaries and OED running on virtual CD drives. Seems like a return to the Windows 3 desktop or even the old IBM translation manager set up.
On the to-CAT-or-not-to CAT issue, I find it much less mental, visual strain to have the segmentation, and it helps me avoid skipping sentences. But, for me, the main time saver is when I get slightly revised versions of the same text, sometimes even a year apart (such as with catalogs updated yearly). Or two versions of a very similar letter (e.s. one to clients, one to suppliers). In those cases my CAT (now Cafe Tran) is priceless. As for productivity, it’s gone down from the beginning. I used to do 2500-3000 and now it’s usually around 2000. I don’t know if I got fussier or what!
Nice article, Corinne, and I have to say I totally agree with your numbers. I usually translate 200-300 words/hour when I’m working on more challenging materials, and 400-600 words/hour in ordinary circumstances.