| Vom Thema belegte Seiten: [1 2] > | A Project with 80% Repetitions? Initiator des Themas: Siddharth Anand
|
|---|
Hi All
I was recently offered a project recently with almost 80% Repetitions
Trados version (6.0, 7.0, 9.0) Percent Words Net
Exact Match
% of base rate for repetitions 25% 818 204.5
% of base rate for 100% matches 25% 119 29.75
Fuzzy matches
% of base rate for 95% - 99% matches 40% 28 11.2
% of base rate for 85% - 94% matches 50% 55 27.5
% of base rate for 75% - 84% matches 60% 54 32.4
% of base rate for 50% - 74% ... See more Hi All
I was recently offered a project recently with almost 80% Repetitions
Trados version (6.0, 7.0, 9.0) Percent Words Net
Exact Match
% of base rate for repetitions 25% 818 204.5
% of base rate for 100% matches 25% 119 29.75
Fuzzy matches
% of base rate for 95% - 99% matches 40% 28 11.2
% of base rate for 85% - 94% matches 50% 55 27.5
% of base rate for 75% - 84% matches 60% 54 32.4
% of base rate for 50% - 74% matches 75% 6 4.5
No Match
% of base rate for No match 100% 43 43
This project totaled 1123 words, but as per the Agency:
1. 818 were Contextual Matches
2. 119 were 100% Repetitions
3. 43 were No Match
Based on this calculation, the 1123 worded document would be charged at 352.85 Words. This literally means only 30% of the Base rate Payment.
Questions:
1. Is this common? I never thought the Repetition could be as much as 70-80% of the source text.
2. How are Contextual matches different from 100% Matches?
3. How to deal with this? Should a Minimum project fee be charged or an increase in base rate
to compensate for this?
Note: This a full comprehensive Text and not a Accounts statement with just numbers. Many of my other clients pay in full (irrespective of the repetitions).
Please help! ▲ Collapse
| | | | Jan Willem van Dormolen (X) 
Niederlande
Local time: 17:03
Englisch > Niederländisch
+ ...
You are lucky, that you have not encountered this before. Clients that pay full rates for matches are actually rather rare.
So the answers to your questions:
1) Yes, this is common. And reps can make up large portions of a text. I once had a job that was 100% 100% matches...
2) As I understand it, 'Contextual matches' are matches that come from the TM, whereas 'Repetitions' are matches that are repetitions of segments found earlier in the same document. Or the other way 'round.... See more You are lucky, that you have not encountered this before. Clients that pay full rates for matches are actually rather rare.
So the answers to your questions:
1) Yes, this is common. And reps can make up large portions of a text. I once had a job that was 100% 100% matches...
2) As I understand it, 'Contextual matches' are matches that come from the TM, whereas 'Repetitions' are matches that are repetitions of segments found earlier in the same document. Or the other way 'round. Something like this anyway.
3) As you are an entrepeneur, you can decide for yourself what you want to do. You could refuse the job unless you are paid in full for all matches (but you will probably not get the job then). You could charge a minimum rate (always a sensible thing to do). You could try and negotiate different rates for the matches, say 50% of the base rate in stead of 30%. It's up to you. There are no 'shoulds' here.
It all depends on how much you want or need the job. ▲ Collapse
| | | | Stanislaw Czech, MCIL CL
Vereinigtes Königreich
Local time: 16:03
Mitglied (2006)
Englisch > Polnisch
+ ...
SITE LOCALIZER | Payment should be adequate to required work | Apr 27, 2011 |
If you translate in Word and don't use any CAT tool it is not an attractive job for you. On the other hand if you use one, it's just a small, easy job and the CAT will do most of the work. As with less than 400 words to translate the job is really tiny, I would advice you to take it and than check how much time you need to accomplish it.
This experience may come very useful when you need to give a quote on a much bigger job of the same kind in the future.
Best Regards... See more If you translate in Word and don't use any CAT tool it is not an attractive job for you. On the other hand if you use one, it's just a small, easy job and the CAT will do most of the work. As with less than 400 words to translate the job is really tiny, I would advice you to take it and than check how much time you need to accomplish it.
This experience may come very useful when you need to give a quote on a much bigger job of the same kind in the future.
Best Regards
Stanislaw ▲ Collapse
| | | | Laurent KRAULAND (X)
Frankreich
Local time: 17:03
Französisch > Deutsch
+ ...
I applied the sliding-scale breakdown an agency sent me to a job I completed for another client.
The result was that my fee would have been cut by some 50%, so that working for this agency on specialised texts at a relatively high rate would have meant being paid less than working on more general texts at a lower rate.
The real question so far with such offers is to determine whether one can reach their hourly rate target or not. IOW would you be able to translate th... See more I applied the sliding-scale breakdown an agency sent me to a job I completed for another client.
The result was that my fee would have been cut by some 50%, so that working for this agency on specialised texts at a relatively high rate would have meant being paid less than working on more general texts at a lower rate.
The real question so far with such offers is to determine whether one can reach their hourly rate target or not. IOW would you be able to translate those 1,123 words in the same time you would need to translate 353 words with the TM provided by the client?
It is a known fact that many agencies will try to get as much work as possible for the lowest possible cost. This is the reason why they use the statistical functions of CAT tools and generally stay with the rate which was initially agreed upon.
So... you know what to do! ▲ Collapse
| | |
|
|
|
Siddharth Anand
Indien
Local time: 20:33
Deutsch > Englisch
+ ...
THEMENSTARTER | Negotiate a Higher Rate and maintain Balanced Grid to ensure that the Hourly Rate is Maintained | Apr 27, 2011 |
Thank you, All!
All the ideas and experiences are very informative. I was able to negotiate a slightly higher rate with the Agency. This should help a bit and as Stanislaw rightly said, this is a small assignment and it will help me learn and decide for big ones in future.
So, I am doing it not for a very good price, but definitely for the experience of it. Let's see how it goes.
Moreover, I think one of the important criteria for Trados grid is to ensure t... See more Thank you, All!
All the ideas and experiences are very informative. I was able to negotiate a slightly higher rate with the Agency. This should help a bit and as Stanislaw rightly said, this is a small assignment and it will help me learn and decide for big ones in future.
So, I am doing it not for a very good price, but definitely for the experience of it. Let's see how it goes.
Moreover, I think one of the important criteria for Trados grid is to ensure that none of the Match Parameters is highly over-weighted. Try to balance the % Rate for each & Match parameter and it will ensure that one is cornered because of the source text being high on any of one or two parameters.
PS. Jan, I do happen to be working with two clients who pay full rate even when TM is supplied for working with Trados. Their only concern is quality and consistency. I guess I am lucky in that case.
I am still learning; there's so much to learn. Thanks again everyone!
Best Regards,
Siddharth Anand ▲ Collapse
| | | | | Context matches | Apr 27, 2011 |
Jan Willem van Dormolen wrote:
2) As I understand it, 'Contextual matches' are matches that come from the TM, whereas 'Repetitions' are matches that are repetitions of segments found earlier in the same document. Or the other way 'round. Something like this anyway.
No.
Repetitions are segments that are repeated within the document but aren't in the TM.
100% matches are segments that occur in the preexisting TM.
Context matches are segments that are 100% matches and are preceded by another 100% match. Context matches (marked by CM in the editor) are more reliable as the segment occurs in the same, well, context.
As to the original question, there is nothing abnormal about it. Some companies have documents that are updated with new data without much change in the actual text. From your perspective, they can be nice cash cows. You'll get a bunch of CM in Trados, which require little reviewing if the translations in the TM are up to scratch. You get paid a 25% rate for reading the source and the translation - whichever way you look at it, that's not bad. You'll get this job done faster than a 350-word job. Of course, conversely, if the TM is rubbish and you spend the rest of your life correcting other people's screwups...
Whether or not you introduce a minimum fee for small jobs is your call and has nothing to do with TM discounts. If you get a lot of jobs like this, a minimum fee would make sense.
[Edited at 2011-04-27 08:32 GMT]
| | | | | Context Match vs. 100% | Apr 27, 2011 |
Actually the difference between a Context Match and a 100% or repetition is simply that with the context match the sentence before and after it are also 100% matches, giving you the idea that the context is the same.
| | | | Attila Piróth
Frankreich
Local time: 17:03
Mitglied
Englisch > Ungarisch
+ ...
| Did you check the quality of the TM? | Apr 27, 2011 |
Before agreeing to any kind of discount (apart from repetitions), you should carefully check the quality of the TM if it contains others' work. You are responsible for the quality of 100% match segments the same way you are for the quality of no match segments: the final version that you submit has to be correct, even if the TM contains errors. It is therefore of the utmost importance to check that the TM is of an acceptable quality, and that a 100% match segment can be carefully revised in the ... See more Before agreeing to any kind of discount (apart from repetitions), you should carefully check the quality of the TM if it contains others' work. You are responsible for the quality of 100% match segments the same way you are for the quality of no match segments: the final version that you submit has to be correct, even if the TM contains errors. It is therefore of the utmost importance to check that the TM is of an acceptable quality, and that a 100% match segment can be carefully revised in the amount of time indicated by your fuzzy scale (in the case you quoted: 25% compared to a no match segment). If the TM is poor, you simply cannot apply the same sliding scale.
In revision, it is very common practice to quote an hourly fee, rather than a per-word price, since the quality of the original translation determines the amount of effort needed by the reviser. Likewise, the quality of the TM can make an enormous difference – from being a very useful tool, which is usually the case when one can use a TM containing only his/her work to a completely useless one when most of the 100% match segments need to be edited. In the latter case all TM discounts should be refused (apart from repetitions).
Finally, checking the TM quality takes some time. This should be treated as an integral part of the quoting procedure: you probably won't charge anything for this, so your prices should be high enough to cover this task as well. Otherwise there won't be any benefits of using the tool for yourself, only for the client.
Best,
Attila ▲ Collapse
| | |
|
|
|
Laurent KRAULAND (X)
Frankreich
Local time: 17:03
Französisch > Deutsch
+ ...
| Only if you don't state something else :) | Apr 27, 2011 |
Attila Piróth wrote:
Before agreeing to any kind of discount (apart from repetitions), you should carefully check the quality of the TM if it contains others' work. You are responsible for the quality of 100% match segments the same way you are for the quality of no match segments: the final version that you submit has to be correct, even if the TM contains errors. It is therefore of the utmost importance to check that the TM is of an acceptable quality, and that a 100% match segment can be carefully revised in the amount of time indicated by your fuzzy scale (in the case you quoted: 25% compared to a no match segment). If the TM is poor, you simply cannot apply the same sliding scale.
(.../...)
I don't want to disagree with you, Attila, but I make the explicit statement that 100% matches will be considered as untranslatables. My take is that agencies are responsible for the legacy TM's they want me to use... and I cannot work for free.
Hence I think it is important not to get nailed down for "one rate fits all" if we are to do additional work like checking legacy TM's or materials of uncertain origins.
| | | | Peter Linton (X)
Local time: 16:03
Schwedisch > Englisch
+ ...
| Can be lucrative | Apr 27, 2011 |
I was once offered a job with 95% repetitions and matches, an update to a user guide that I had previously translated. So almost everything had already been approved both by the customer myself.
The customer demanded a large discount, which I was happy to give, because the translation took much less time than usual. So although the rate per word was ridiculously low, the rate per hour was very good – in fact double my normal hourly rate. A good win-win situation.
| | | | Attila Piróth
Frankreich
Local time: 17:03
Mitglied
Englisch > Ungarisch
+ ...
| Who takes the responsibility? | Apr 27, 2011 |
Laurent KRAULAND wrote:
I don't want to disagree with you, Attila, but I make the explicit statement that 100% matches will be considered as untranslatables. My take is that agencies are responsible for the legacy TM's they want me to use... and I cannot work for free.
That's a possible approach - 100% matches are simply excluded: no money, no responsibility. A case can be made for this approach for context matches, where not only the sentence is identical to a TU in the TM but also the one preceding it, as is usually the case when a whole paragraph is taken over from a legacy material. Many CAT tools offer this separate category, and exclude, by default, context match segments from the translatable material.
But extending this approach to simple 100% matches is another matter. If the whole text is revised by a second person, this approach can still be applicable - after all, a professional takes the responsibility of the correctness of the translation. But if the translation of the different segments in a single paragraph may be the patchwork of 10 different translators, without a final revision, then there are at least two serious issues:
1.) The translation will be perfectly inconsistent in style. I have seen that more than once: in instruction manuals the imperative ("Press the red button") translated in at least three different ways (singular "you", plural "you", "we" ["let's press the red button"): each of which is an option but not a mix of the three.
You may say this is not your problem. Right - but this is tantamount to saying that the stylistic part of your work can be separated from the rest. In other words, that you do not consider style as an important differentiating factor that could put you above your peers. By taking part in a project with many translators, you agree to provide a service that does not make use of your full skill set. Will you be able to impose your conditions better than some of your competitors who may not even possess the skills that you dismissed so easily?
This is the core problem with low-profile translation jobs, first and foremost with MT post-editing, where the name of the game is "productivity", and the speed benchmark is set so high that you simply cannot perform the same necessary research that you are used to. So, you rely on TM context research, and put "consistency with legacy material" above "correctness". The moment you agree to dumb down your services, you will be in competition with far too many others, and your expertise and previous experience will lose all their value.
Disclaimer: These are just the thoughts that came to my mind about your comment, Laurent. In the previous paragraphs, "you" is a generic "you", and I would never imply that you personally work that way.
2.) Segments that come from completely different contexts may be completely wrong, even if the subject matter of the TM is restricted. Here is an example I encountered: in the description of the connection of a urinary catheter, "Male" appears as a single segment in a figure. The TM contains a hit, which means "the male part of a male-female connection". But here the meaning was "for male patients". (In many languages, these two meanings of "male" may be translated by the same word, so the issue does not even exist. But this makes the situation no less dangerous: a PM speaking in addition to English, say, French, where "mâle" works both ways, may assume that there is no issue here at all.)
You may say this is an exceptional coincidence. Well, somebody has to vouch for the absence of such exceptions in the final translation. And when the translation is put together from different parts, without at least one professional checking the whole thing, such exceptions simply won't be excluded.
But I guess this is not the case for Siddharth, as 100% matches are paid to some extent. And if they are paid, he will bear full responsibility for their correctness.
Best,
Attila
| | | | | The older and easier solution: Compare documents in Word. No CAT, no greed, no anger, no cry! | Apr 27, 2011 |
Sometimes it really worries me, how "CAT blind" we translators already are.
Oh my dear, in cases with such high percentages of 100% matches, context matches and repetitions (updated version of an earlier translated document), wouldn't it be easier to do it the "old fashioned way" and to use common sense? I mean to compare the two versions of the source text, to take the translation of the older version, to copy in the updated parts from the updated version of the source text and to ... See more Sometimes it really worries me, how "CAT blind" we translators already are.
Oh my dear, in cases with such high percentages of 100% matches, context matches and repetitions (updated version of an earlier translated document), wouldn't it be easier to do it the "old fashioned way" and to use common sense? I mean to compare the two versions of the source text, to take the translation of the older version, to copy in the updated parts from the updated version of the source text and to translate only these parts (with the CAT tool, if useful)?
This way we really won't have to worry about mistakes the CAT tool could introduce in the repeated parts or the matches (numbers, formatting and so on), when the Trados grid doesn't allow us to have a closer look at them.
I was doing it like that all the time before the Trados greed came up and I invoiced only the newly translated parts plus some amount of time for the compare task and to get familiar with the terminology.
Kind regards,
René Stranz-Nikitin
www.uersn.de ▲ Collapse
| | |
|
|
|
Jaroslaw Michalak
Polen
Local time: 17:03
Mitglied (2004)
Englisch > Polnisch
SITE LOCALIZER
René Stranz-Nikitin wrote:
Wouldn't it be easier to do it the "old fashioned way" and to use common sense? I mean to compare the two versions of the source text, to take the translation of the older version, to copy in the updated parts from the updated version of the source text and to translate only these parts (with the CAT tool, if useful)?
Easier? Debatable. Faster? No way.
| | | | | MS Word is comparing documents very fastly. | Apr 27, 2011 |
Jabberwock wrote: René Stranz-Nikitin wrote:
Wouldn't it be easier to do it the "old fashioned way" and to use common sense? I mean to compare the two versions of the source text, to take the translation of the older version, to copy in the updated parts from the updated version of the source text and to translate only these parts (with the CAT tool, if useful)?
Easier? Debatable. Faster? No way.
Hi Jabberwock,
My impression is that you already forgot how fast MS Word compares documents.
But I know, we are just belonging to different camps: You are a fan of the table-like designed CATs (Déjà Vu, Studio, MemoQ) and I like to work in Word+TWB.
René
[Edited at 2011-04-27 12:17 GMT]
| | | | Jaroslaw Michalak
Polen
Local time: 17:03
Mitglied (2004)
Englisch > Polnisch
SITE LOCALIZER | Oh, I remember... | Apr 27, 2011 |
René Stranz-Nikitin wrote:
My impression is that you already forgot how fast MS Word compares documents.
I remember it quite well: I have used the method you described for years when I updated translations of digital camera manuals (when I could not even afford one yet!).
The speed of comparison is, of course, negligible. It's the jumping, copying, switching and pasting that is time consuming...
But indeed, I had in mind the tabular CAT interface, where I can jump to/filter the relevant segments instantly. Revising a bilingual file in Word would not be so easy, so with the project of that length your method might work better (unless there are many formatting changes).
| | | | | Vom Thema belegte Seiten: [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 » A Project with 80% Repetitions? | Wordfast Pro | Translation Memory Software for Any Platform
Exclusive discount for ProZ.com users!
Save over 13% when purchasing Wordfast Pro through ProZ.com. Wordfast is the world's #1 provider of platform-independent Translation Memory software. Consistently ranked the most user-friendly and highest value
Buy now! » |
| | Trados Studio 2022 Freelance | The leading translation software used by over 270,000 translators.
Designed with your feedback in mind, Trados Studio 2022 delivers an unrivalled, powerful desktop
and cloud solution, empowering you to work in the most efficient and cost-effective way.
More info » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |
Login to reply/comment 
Translation news
