https://www.proz.com/forum/money_matters/210899-repeats_paid_or_not.html

Pages in topic:   [1 2] >
Repeats: Paid or not?
Thread poster: erickl

erickl
New Zealand
Local time: 21:41
English to French
+ ...
Oct 30, 2011

As a translation agency, and like many others in the business, MSDS-Translations has reviewed some of the current tendencies of the market, one of them involving the use of Translation Software and the increasing tendency of pressuring the translators to accept to process the repeats at no charge.
We have been trying to be fair in the process and to look at both the translation agency and the translator's side. If one is seriously in the translation business for the long run and not in the
... See more
As a translation agency, and like many others in the business, MSDS-Translations has reviewed some of the current tendencies of the market, one of them involving the use of Translation Software and the increasing tendency of pressuring the translators to accept to process the repeats at no charge.
We have been trying to be fair in the process and to look at both the translation agency and the translator's side. If one is seriously in the translation business for the long run and not in the business of making as much money as possible in the shortest time, fairness has to be applied to both customers and suppliers, i.e., the translators, and especially the freelance translators.

What follows is a summary of our analysis.
Could we do without CAT tolls? The answer to this was quite simple in our segment of the market, and this has to be read in the prospect of a company dealing with documents which are very repetitive by nature. Our not using CAT Tools would simply be unsustainable. Yet, there is a fact that did not raise even the slightest doubt in our mind at the end of our process: CAT Tools are designed for the benefit of translations agencies first, client second, and the translator unfortunately seems to be of very little concern.
Some of these tools can add several hundred tags in a single document, and although the translator has to deal with them and eventually find out that they can at times be more work than the words themselves, they ignore them in the word count. The same phrase with 5 tags is considered a 99% repeat in some cases, when in fact, the translator has to rebuild the entire segment.
Others add a very minimum of tags, and they are per say negligible, but this is the exception rather than the rule. In any case, we found the tags an issue that is very arguable when it comes to the counting of the repeats and fuzzy segments.

The other problem we see is that a repeat in the source language, even on a single word, might not be a repeat in the target language. Any adjective (as simple as good) would have several different forms in the languages which have an adjective masculine, feminine, and possibly neutral forms. ("bon", "bonne" in French). Place this adjective within a phrase and it might turn out to be quite different. The same goes for the verbs and many other particularities which are language-specific.
The bottom line is that repeats are actually based on what the customer and the translation agency see, with total disregard for the work the translator might actually have to do. The assumption that a translator just has to click a button for every repeat, conveniently adopted by the translation agencies, is in any case far from reality.

Within this view, it would more fair if the repeats were analyzed on the translated file in the target language rather than on the source file, a practice that is adopted in some languages for which the translators charge on the target language word count, but again, this is not the general rule and we do not feel that it would realistically be doable for us to implement rules that are not in accordance with business common practices.

Once the usage of the CAT Tools is taken as a must-have, the next question for us was: Which one should we go for?

We came to the conclusion that, unless we would decide to go on a server-based implementation, a step we are not considering in this paper or at this time, we really had very little to say.

This might come as a surprise, but in fact, it sounds to us quite pretentious to demand a quality translation and to tell the translator on the other hand: "this is what you are going to use, and this is how you are going to do it". We have a responsibility as a translation agency; on their side, the translators also have to assume theirs. To leave with them the tool of their choice, for economical, personal reason or simply because it is the tool they like best or are more used to seems to us to be the only reasonable choice.

We do use translation memories and reuse them, and we feel it is a service we can require from a translator. We can accommodate several formats, and any decent tool should, at the very least, be able to extract a memory in standard memory exchange format. We feel that this and should be perfectly acceptable for anybody who claims to be in the translation business.

Any translator has the right not to offer this service, and we perfectly respect this, it simply is in our view a business decision each party has to make, and ours, in the scope of our market, is to make use of it for most documents.

Our legal contacts also raised another aspect of the problem we have not fully investigated inasmuch as it became irrelevant in our case, and which I will leave to somebody more versed in legal questions: If a translator is not being paid for the repeats on a document, he nevertheless incurs costs to process this document and these repeats (if only to get an internet connection or a computer). Therefore, this document is at least partially processed at his own cost, which places a serious doubt of the validity of any confidentiality agreement between the translator and the translation agency, no matter what this agreement might stipulate.

Looking to the future, can we ignore the tendency of not paying for repeats?

Once more, we look at our situation and our range of the market, and we do not pretend to present a general solution for the whole industry. We are running a business involving highly technical and specialized documents, and we are certainly not claiming to address all aspects of the industry.

Our conclusions are:

1. CAT tools are a feature we absolutely need to be able to use, and we have to remain in a position allowing the usage of the translator's tool of choice. Also, the most known and/or most expensive software are not necessarily, in our view, the best choice, but it is a choice we can make for ourselves and not for independent contractors.
2. Repeats can give more work to the translator than generally acknowledged, and eventually turn out not be repeats at all in their working language. Ignoring this fact to make a case for the non-payment of them is unfair to the translators and at the very least on the border of being unethical.
3. A legal question has been raised as far as not paying the repeats, a subject we have not further investigated, inasmuch as the outcome of the analysis would not have changed our decision.
4. The claim that CAT-Tools lead to a poorer quality is in our view not justified. Like any other tool, it is only as good as one uses it. Good translators who use them well make no more mistakes with CAT Tools than without them.

The future will tell whether we are right or wrong, and we do not pretend to be beholder of the truth; we are convinced on our side that our decision of not asking the translators for free repeats is right at this point of time and in the foreseeable future, for our segment of the market.
Ultimately, the translators have the last word; if we cannot survive as a translation agency by being out of sync with the market, we can no better survive without very skilled and specialized translators. They, on their side, have to consider as we do the segment of the market they are in or work for. As painful as it might be when it happens, not caving in on any particular rate-reduction demand might turn out to be more profitable in the long run.
The translation business as a whole is changing fast; the tendency is for some even reputable companies to follow the pressure created by those who try to cut the price by any means. But it would be wrong in our opinion to apply this general rule to specialized areas of the market, in which the quality of the translation and of the translator is paramount. No matter which business we are into, there will always be somebody who claims to be able to do it cheaper. It is cheaper as long as quality and customer service remain the same, and customers are very rarely fooled more than once.

But we do regret that we are stuck in a position in which we have to argue against the quick-sale and somewhat misleading marketing strategies of the CAT Tools software providers on one hand, the needs and sometimes unrealistic expectations of some clients of the others, all the while trying to remain fair to the people who are the core of our business.

The question can be put in a very brutal manner? Do we put our head in the sand, ignore reality, and blindly accept the marketing claims of the software providers at the cost of our translators, or do we hold our ground and remain a neutral party seeking to be fair to both sides?

We happen to believe that if we are fair to both translators and customers, they will remain loyal to us. Translation has traditionally be an art as much as a profession and we are convinced, without by any means rejecting technological improvements, that it can, should and will remain as such. If we look at it in a long-term perspective, there is no other way to go. This is the only perspective we are interested in.

Erick.
MSDS-Translations Ltd.
Collapse


 

Tomás Cano Binder, BA, CT  Identity Verified
Spain
Local time: 11:41
Member (2005)
English to Spanish
+ ...
Pay for work that is done, do not pay for work not done Oct 30, 2011

In the well-known Spanish play Don Juan Tenorio (traditionally played in the night of All Saints' day, i.e. next Monday), the innkeeper of the Hostería del Laurel, where the play starts, is paid by two masked guests who just want to be left alone in a table and do not want to be served anything. He claims in sheer joy that these are the best guests: they pay for what they do not consume, and there's a good way to prosper!

As a technical translator who has been using CAT tool
... See more
In the well-known Spanish play Don Juan Tenorio (traditionally played in the night of All Saints' day, i.e. next Monday), the innkeeper of the Hostería del Laurel, where the play starts, is paid by two masked guests who just want to be left alone in a table and do not want to be served anything. He claims in sheer joy that these are the best guests: they pay for what they do not consume, and there's a good way to prosper!

As a technical translator who has been using CAT tools for well over a decade, all I can say is that I like to be paid for work I do, and dislike to be paid for work I do not do. So while I like to be paid for repetitions and high matches, I think it is fair that I get paid only part of the full rate for them.

On an average, I ask my customers to pay about 20%-25% of the full rate for repetitions, and the full rate for anything under 75% of match. I think this arrangement correctly compensates me for my investment/knowledge of the CAT tool and the actual work involved in the segments. It is also fair to the end customer, and keeps nice work rolling in.

It strikes me that a technical translation firm still exists in 2011 which does not use CAT tools, among a skilled group of competitors using CAT tools and effectively selling their value. You must have a nice niche market up the sleeve!

[Edited at 2011-10-30 07:38 GMT]
Collapse


 

erickl
New Zealand
Local time: 21:41
English to French
+ ...
TOPIC STARTER
Not so special Oct 30, 2011

Hi Tomas,

Thanks for your comments, and sorry if the post is somewhat misleading, but we do use CAT tools and TMs, the point for us was to reassess our use of the CAT Tools and whether we should or not settle on one of them as well as to re-evaluate our handling of the repeats. So we are not so special after all...


 

Nicole Schnell  Identity Verified
United States
Local time: 02:41
English to German
+ ...
In memoriam
My take Oct 30, 2011

Here is how my clients and I are dealing with this issue:

100% repetitions are paid at my proofreading rate, i.e. 30%.

"Fuzzies" are paid at the full rate.

No third party TMs are accepted.


I don't even own any CAT tools, the editor versions of the desired CAT tools are provided by the client.

Why I don't accept any discounts for fuzzy matches:
A large portion of my technical translations involve chemical formulas.
... See more
Here is how my clients and I are dealing with this issue:

100% repetitions are paid at my proofreading rate, i.e. 30%.

"Fuzzies" are paid at the full rate.

No third party TMs are accepted.


I don't even own any CAT tools, the editor versions of the desired CAT tools are provided by the client.

Why I don't accept any discounts for fuzzy matches:
A large portion of my technical translations involve chemical formulas. Which by nature involves tons of annoying and time-consuming tags. There is no such thing as the CAT tool being a time-saver if the formulas have to be rearranged in a different order for each product description. Carving them in wood would be faster than arranging single letters and numbers around tags, yet this is considered a fuzzy match that is supposed to come with a hefty discount. I fail to see any logic in this kind of calculation.
Collapse


 

Tomás Cano Binder, BA, CT  Identity Verified
Spain
Local time: 11:41
Member (2005)
English to Spanish
+ ...
A quick reply Oct 30, 2011

erickl wrote:
Thanks for your comments, and sorry if the post is somewhat misleading, but we do use CAT tools and TMs, the point for us was to reassess our use of the CAT Tools and whether we should or not settle on one of them as well as to re-evaluate our handling of the repeats. So we are not so special after all...

Well, if this is the case, clearly you should settle for a CAT tool capable of managing/outsourcing/receiving files in standard formats, like XLIFF for bilingual files and TMX for memories. Being able to handle bilingual Trados-compatible Microsoft Word files would be an advantage.

If you want to secure good translators you need to give them some leeway in the choice of a CAT tool, so there are options that are not generally accepted, like having to translate online in your systems. We really prefer to receive files and deliver files, rather than having to work over the Internet with a system we did not choose.


 

neilmac  Identity Verified
Spain
Local time: 11:41
Spanish to English
+ ...
Exactly Oct 30, 2011

erickl wrote:

2. Repeats can give more work to the translator than generally acknowledged, and eventually turn out not be repeats at all in their working language. Ignoring this fact to make a case for the non-payment of them is unfair to the translators and at the very least on the border of being unethical.

Erick.
MSDS-Translations Ltd.


My thoughts exactly. Since my base rate is average-low, I don't usually do discounts for repeats but last week I gave one client an unsolicited 5000 word rebate on a project because he's a quick payer...


 

Samuel Murray  Identity Verified
Netherlands
Local time: 11:41
Member (2006)
English to Afrikaans
+ ...
My comments Oct 30, 2011

erickl wrote:
The same phrase with 5 tags is considered a 99% repeat in some cases, when in fact, the translator has to rebuild the entire segment.


True, but I think the idea is that you win some and you lose some. Just as it is not generally seen as unfair to charge the same rate for long words and for short words (because the long and short words tend to average each other out), so too (in an ideal world) multitag segments will average out between those that require a lot of work and those that require no work.

A leveraging system that takes this into account, however, could be useful (but how to design it and keep it simple, is the question). For short segments, one could say 1 tag = 2 words, and take that into account when doing leveraging (and rate calculations). For long segments, one could say that for each tag in a tag:word ratio of 1:10 the match percentage is penalised by 1% (added to existing penalties). Sounds very nice in theory, but who decides the ratios, and how scientific are they really?

The other problem we see is that a repeat in the source language, even on a single word, might not be a repeat in the target language.


A good (partial) solution to this is in-context repeats matching (so-called 101% matching) that requires a context of not just 3 segments but more something like 10 segments. In this scenario, the translator will then be paid only for repetitions that occur reasonably close to non-repeating text.

This might come as a surprise, but in fact, it sounds to us quite pretentious to demand a quality translation and to tell the translator on the other hand: "this is what you are going to use, and this is how you are going to do it".


There is something to be said for having all translators use the same tool -- if any of them have a problem, the PM will be more likely to recognise it. Otherwise, you expect the PM to be an expert in all CAT tools. If a translator fails to deliver (either too late or not in the right format) because of some incompatibility in his CAT tool, then it doesn't matter whose fault it is, but the agency suffers.

If the translation is late or if the delivered file is broken, it is easy to spot, but if the CAT tool simply broke the translation in a non-obvious way, the fault may only be discovered by the time the end-client has spent a lot of money already.

Or, a translator may tell the client "my CAT tool is fully compatible with X" because the web site says so or because the user group says so, but if it turns out that there is an incompatibility (in segmentation, in workflow, or in delivery format), the agency suffers.

If a translator is not being paid for the repeats on a document, he nevertheless incurs costs to process this document and these repeats (if only to get an internet connection or a computer). Therefore, this document is at least partially processed at his own cost, which places a serious doubt of the validity of any confidentiality agreement between the translator and the translation agency, no matter what this agreement might stipulate.


I don't understand this issue.

The translator is paid for the job as a whole, and the word count is simply a comfortable way to determine the cost of the job. The translator isn't paid for the individual words, but for the file, or the job, or the project, and to calculate the estimated time of the job, he uses the [weighted] word count, which is simple, effective, and generally fair.

2. Repeats can give more work to the translator than generally acknowledged, and eventually turn out not be repeats at all in their working language.


How about to two-tiered quoting system that includes leeway for "unforeseen" circumstances?

I mean, the quote would set the minimum price of the job at the assumption that all non-101% repeats are exact repeats, and an accompanying maximum price which assumes that all non-101% repeats had to be edited.

Then, after delivery, the translation is analysed to determine which non-101% repeats were edited by the translator. The translator then gets a low payment for non-edited non-101% repeats and a standard per-word proofreading payment for edited non-101% repeats.

In this way, the translator is compensated more fairly, but the client still knows what the maximum price is that he could potentially pay.

The question can be put in a very brutal manner? Do we put our head in the sand, ignore reality, and blindly accept the marketing claims of the software providers at the cost of our translators...


The problem is that if agencies expect translators to do more and more work in the same amount of time (based on the leverage analyses only), those translators will be forced to deliver substandard work (it may not come suddenly -- it may occur gradually).


 

Anne Seerup
Ireland
Local time: 10:41
English to Danish
+ ...
Not beneficial to us at all... Oct 30, 2011

Most of the time I find these Tageditor jobs that involve huge client memories exceedingly time-consuming. I also find that it is really easy to miss fuzzy matches when my eyes are distracted by all the annoying tags. Not to mention the eye strain of looking at that horrendous tagged text with no option to view and proofread the finished layout.
And when reusing material translated by 50 different people in the past the consistency goes right out the window, and on top of that the transla
... See more
Most of the time I find these Tageditor jobs that involve huge client memories exceedingly time-consuming. I also find that it is really easy to miss fuzzy matches when my eyes are distracted by all the annoying tags. Not to mention the eye strain of looking at that horrendous tagged text with no option to view and proofread the finished layout.
And when reusing material translated by 50 different people in the past the consistency goes right out the window, and on top of that the translator is often made responsible for TM errors anyway, because nobody knows who did what.
Whenever there has been a problem it has involved one of these jobs. The client then creates a PDF with the final layout, which is given to some anonymous client reviewer, who doesn't have a clue who did what, and just marks everything they spot in the file, including layout errors. I have had to spend ages tracking down the errors marked in a PDF to see if they were even mine, or just something from the TM, that I was told to adhere to. Did I get paid for this? Of course not. Well not, unless I spent hours of my precious time proving that a large percentage of errors were from the memory.
I guess my point is here: We are still responsible for the end result, regardless of repeats and matches.

Of course repeats should be paid - we after all need to insert tags, revise the segments, check the for consistency if we make changes in one. They can be quite time-consuming if there are lots of them.

So no, CAT tools do NOT save time, please just scrape away that rosy image conveyed by the industry and put it in the bin where it belongs. Why on earth should we do anything at no charge? I wonder how the work would get done without a translator
Collapse


 

Heinrich Pesch  Identity Verified
Finland
Local time: 12:41
Member (2003)
Finnish to German
+ ...
I don't look at reps if I'm not paid Oct 30, 2011

For quality translation it is true, that reps need attention. But if the customer chooses not to pay for it, it is at their responsibility, not mine. In such cases I filter the less than 100% matches and translate and proofread only these once.

 

Russell Jones  Identity Verified
United Kingdom
Local time: 10:41
Italian to English
Thank you Oct 30, 2011

Thank you Erick for sharing this analysis and for your thought-provoking comments.

 

Helmaninquiel
Local time: 11:41
English to Spanish
+ ...
You are so right Oct 30, 2011

Anne Seerup wrote:


Most of the time I find these Tageditor jobs that involve huge client memories exceedingly time-consuming. I also find that it is really easy to miss fuzzy matches when my eyes are distracted by all the annoying tags. Not to mention the eye strain of looking at that horrendous tagged text with no option to view and proofread the finished layout.
And when reusing material translated by 50 different people in the past the consistency goes right out the window, and on top of that the translator is often made responsible for TM errors anyway, because nobody knows who did what.
Whenever there has been a problem it has involved one of these jobs. The client then creates a PDF with the final layout, which is given to some anonymous client reviewer, who doesn't have a clue who did what, and just marks everything they spot in the file, including layout errors. I have had to spend ages tracking down the errors marked in a PDF to see if they were even mine, or just something from the TM, that I was told to adhere to. Did I get paid for this? Of course not. Well not, unless I spent hours of my precious time proving that a large percentage of errors were from the memory.
I guess my point is here: We are still responsible for the end result, regardless of repeats and matches.

Of course repeats should be paid - we after all need to insert tags, revise the segments, check the for consistency if we make changes in one. They can be quite time-consuming if there are lots of them.

So no, CAT tools do NOT save time, please just scrape away that rosy image conveyed by the industry and put it in the bin where it belongs. Why on earth should we do anything at no charge? I wonder how the work would get done without a translator




CAT tool may save time on the one side, but even in some 100% matches translator may have to add tags, format something...
Apparently agencies don't take that into account.
If we must apply discounts for fuzzy matches, we also should charge for this "extra" work.
There's no doubt CAT tools are useful, but not the panacea.


 

erickl
New Zealand
Local time: 21:41
English to French
+ ...
TOPIC STARTER
Comments Oct 31, 2011

Samuel Murray wrote:
True, but I think the idea is that you win some and you lose some.


I fail to see how one can win some on repeats that are not being paid.



There is something to be said for having all translators use the same tool -- if any of them have a problem, the PM will be more likely to recognise it. Otherwise, you expect the PM to be an expert in all CAT tools. If a translator fails to deliver (either too late or not in the right format) because of some incompatibility in his CAT tool, then it doesn't matter whose fault it is, but the agency suffers.


It is for the agency to assume the responsibility of managing a project and to adequately select its translators, and if it fails to do so, it can only blame itself if it suffers. As for having all translators using the same tools, we have indeed looked at it very seriously, and our conclusion is that the most efficient tool is the TMX, NOT the software itself. If our deliverables are the translated file and the TMX, both Project managers and Translators can use the tool they are the most comfortable with, and even a PM who knows one single tool is able to handle it.


The translator is paid for the job as a whole, and the word count is simply a comfortable way to determine the cost of the job.


In our world, quotes and most POs are by word, often stating fuzzy: xx%, repeats, 0% and not by file or project, with the exception of small jobs falling into the minimum fee-category. I cannot think of one single translator who does not give us quotes based on the word count. We are looking at what we face, not the interpretation of it. The validity of a non-concurrence clause pertinent to services, according to our information, is highly questionable if part of the services provided is not remunerated, and at the cost of the provider. Once more, we have not gone deeply in the matter, as it turned out too be irrelvant in our case



How about to two-tiered quoting system that includes leeway for "unforeseen" circumstances?


I believe this would inevitably lead to giving the customer a "safe price" quote, with the risk of missing the order; hardly an advantage for us or for those who work for us.


The problem is that if agencies expect translators to do more and more work in the same amount of time (based on the leverage analyses only), those translators will be forced to deliver substandard work (it may not come suddenly -- it may occur gradually).


This is part of the broader challenge as we see it: to remain competitive, and be able to increase productivity without sacrificing quality. The easy and too common solution is to demand more from the translators, we believe that, as a translation agency, we also have to do our bit.


 

erickl
New Zealand
Local time: 21:41
English to French
+ ...
TOPIC STARTER
Appreciated Oct 31, 2011

Russell Jones wrote:

Thank you Erick for sharing this analysis and for your thought-provoking comments.


Thanks Russell


 

Graça Ribeiro  Identity Verified
Local time: 10:41
English to Portuguese
+ ...
Client should pay the service requested Oct 31, 2011

Heinrich Pesch wrote:

For quality translation it is true, that reps need attention. But if the customer chooses not to pay for it, it is at their responsibility, not mine. In such cases I filter the less than 100% matches and translate and proofread only these once.


If checking/reviewing of 100% matches and repetitions is requested by the client, then a (lower) fee must be paid for that. If the client agrees that the translator totally ignores 100%/reps, then no payment is due.

Sorry for the over-simplified reply/comment to an extremely complex question/matter.


 

Helmaninquiel
Local time: 11:41
English to Spanish
+ ...
Not always 100% are 100% Oct 31, 2011

Graça Ribeiro wrote:

If checking/reviewing of 100% matches and repetitions is requested by the client, then a (lower) fee must be paid for that. If the client agrees that the translator totally ignores 100%/reps, then no payment is due.

Sorry for the over-simplified reply/comment to an extremely complex question/matter.


It all depends on the quality of the translation memory. Sometimes due to segmentation, a 100% match can be totally different. So what should we do in that case?


 
Pages in topic:   [1 2] >


To report site rules violations or get help, contact a site moderator:


You can also contact site staff by submitting a support request »

Repeats: Paid or not?

Advanced search


Translation news





Protemos translation business management system
Create your account in minutes, and start working! 3-month trial for agencies, and free for freelancers!

The system lets you keep client/vendor database, with contacts and rates, manage projects and assign jobs to vendors, issue invoices, track payments, store and manage project files, generate business reports on turnover profit per client/manager etc.

More info »
SDL Trados Studio 2019 Freelance
The leading translation software used by over 250,000 translators.

SDL Trados Studio 2019 has evolved to bring translators a brand new experience. Designed with user experience at its core, Studio 2019 transforms how new users get up and running, helps experienced users make the most of the powerful features.

More info »