Cat tools, globalization and the big money grab
Thread poster: Martin Wenzel

Martin Wenzel
Local time: 13:57
English to German
+ ...
Jun 22, 2007

Two domineering GIM (explained below) providers join, I am wondering what future implications this trend will have on translators...

Here is an interesting question from the FAQs of the ZZZZZ website:

What is the joint vision of XXX and YYYYY?
Answer: Both XXXX and YYYY share the common vision that to compete in a truly global world, companies must adopt a Global Information Management (GIM) strategy to demonstrate accelerated time-to-market with higher consistency of global content. Software content is a critical component of GIM, extending GIM capability to product lifecycle management.

Well, well, who is competing with whom? We, the translators, who buy Cat tools in order to make agencies richer, who cut down on our rates because of the repetitions they don't want to pay, which, in most cases [unless the final customer is a strong believer in CAT tools, i.e. knows about their existence and that they might save him a pretty penny) however, will be charged to the final customers...

I really feel translators are loosing out in the big money grab game...

Perhaps we should go on a CAT tool strike, i.e. still use these tools, but refuse lower payment thanks to repetitions...

Yes, I am using XXXXXX, but I am sorry to tell you this will not be reflected in your invoice...

I am a strong believer in CAT tools, but not to make agencies or CAT tool creators richer, who cannot even write proper manuals, let alone explain how their software works...

Direct link Reply with quote
You hit the nail right on the head (of my mind)! Jun 22, 2007

Martin Wenzel wrote:

I really feel translators are loosing out in the big money grab game...

Perhaps we should go on a CAT tool strike, i.e. still use these tools, but refuse lower payment thanks to repetitions...

--------- omission --------------

I am a strong believer in CAT tools, but not to make agencies or CAT tool creators richer, who cannot even write proper manuals, let alone explain how their software works...

Thank you very much for speaking my thoughts!
When I started this business of translation, I collected 0.35 USD per word. It was nearly 20 years ago. How much is it now?

Recently someone inquired my availability, said they are impressed by my background. Sounds good, hah?
Then they asked my CAT discount rate as I say I have/utilize XXXXX. I said I mostly not offer such discount, or 85% and above repetitions only.
They ceased to communicate with me.

Who came up with this demeaning, insane business of repetition discount anyway? CAT tool makers or Agencies?
It is my strong belief that it is them who should provide us CAT tool. They should develop their proprietory CAT tool themselves. Ally with other agencies if they by themselve cannot afford such expenses. Or for those translators who already own them, subsidize in whateve way both translator and the agency agree, instead of ripping off us further.

Somewhere in this forum in different thread, there were discussions about discount rate. Then the issue was where to draw the line.
We know very well, even for same word, same translation does not work in different context, which often is the case especially in big documents. We have to devise appropriate ones for second, third, or fourth repetitions. In extreme case, they all could be different picks. This cost us brain and time, yet to the eyes of these stupid "word analysis", there is no distinction. Some agency even suggest 0 pay for 100% repetition. How mindless! How insulting!
Are you so willing to take this nonsense?

Yes I love CAT tools. Jobs become easier with them. But that's the convenience I paid for. Would I have to get paid less simply because I have DTP tool, for instance?

I am with you Martin. I will sign up for you.

Direct link Reply with quote

Boris Sigalov
Local time: 14:57
English to Russian
No CAT tool discounts Jun 23, 2007

Martin Wenzel wrote:

Perhaps we should go on a CAT tool strike, i.e. still use these tools, but refuse lower payment thanks to repetitions...

Martin, for me it's not a strike but normal work flow: I simply never offer any CAT tool discounts.

Direct link Reply with quote
István Lengyel
Local time: 13:57
English to Hungarian
+ ...
discounts vs. productivity gains - the vendor's dilemma Jun 24, 2007

I hope you don't mind if I tell you the CAT tool vendor's dilemma. I am not from XXXX but from ZZZZ, and we had a very close cooperation with YYYY before it became XXXX (confusing, eh?:)). So the basic idea behind any statistical analysis (repetitions, similarities, internal similarities and one could also add items such as term coverage, average length of segments in words, etc.) is to estimate the productivity gain from using a translation environment. But what's productivity? How can it be measured? I believe the only way to measure productivity is time - how much time does it take for you to translate a project. But of course there are translators who translate fast, others translate slower. So maybe you could have a productivity difference percentage, like 'on average people would translate this project 40% faster with the use of a tool than without it'. But after all, productivity is a personal category.

The problem is, however, that it's not really the translation memory that saves you time. You have three major 'translation categories' - please forgive me my simplistic model:

1. reference and research, while you are looking for information,
2. the formulation of the translation, while you are thinking about how to translate,
3. the entry of the translation, while you are typing it and formatting it.

Of course all three can coincide in time, but I believe these are the three major items.

As regards #1, I believe the term base is more important than the translation memory because in the case of long sentences even with color-coding the checking and correction of differences between the source segment and the memory segment can take long. As regards #2, a TM may even slow down your thinking because you have to adapt sentences, not formulate some. As regards #3, nothing is more important than a good editing environment, one that is intuitive and helps you with the typing/editing.

So I believe that statistics can only cover a small proportion of the actual productivity gain. And the problem is that there is no way to train this. Statistics takes into account only the text, not the translator.

I don't know how the regular numbers (like 30% for repetition, slightly more for good matches, etc.) evolved, but probably there was no experimentation behind this. And I don't think anyone ever timed his or her translations with or without translation tools. I think such an experiment would be great.

Statistics can unravel information about the text. The state of the art is pretty simple. In ZZZZ we introduced the concept of homogeneity, which also gives you fuzzy matches even if you don't have a TM, based on simulating the building of a translation memory. There are lots of other things that can help in the estimation. I believe that the length of a sentence increases complexity, and it takes less time to translate 10 sentences with 6 words in each than 5 sentences with 12 words in each. Beyond any doubt, statistics tell you something valuable.

The problem is, though, that this information is available to everyone. Translators, translation companies, end clients. And there is a chain: clients fight with translation companies, translation companies fight with translators to get more and more discounts. When we come up with a new idea (and I think originally this was the case of XXXX too) we believe that it should indicate something, a piece of information worth considering. If people press for lower prices, you will be able to see if it's worth accepting it. Suppose you work for 10 cents a word normally. Then you get a very simple job with a high coverage and the client wants it for 8 cents a word and you take it. No CAT discounts yet. Then the regular CAT discounts come into play and the price of translations goes down. I would personally only care if I make less in a day than before. But it's really hard to tell whom it is good to have detailed statistics.

XXXX has been historically concentrating on the translation end-user and so do all other successful vendors (well, I believe there is only one other so far). End-users have a good chance to save money. They do everything to save money, ask for discounts from the translation company and the translation company does not want to take discount risks.

I worked for a natural language processing company before and we had a very special job, translating trademark descriptions from Central-European languages into English. It was a custom web-based system, in a way primitive but enhanced with simple machine translation. The real problem, however, was that the client wanted to have the record as a unit. And a record can be anywhere between 50 and 5000 characters... So in this case you can either add a 'puffer' to offset potential losses or ask the translators to count in records too. We did the puffer thing, and as a consequence we did not make any profit as the average length of records increased later...

I just wrote down this story because this is something that explains why a translation company is pressing for discounts too.

So, for a translation environment vendor it is really hard to tell what statistical numbers are good: those that show a lot of things or those that barely show anything. If we go for telling a lot, that's welcome by the translators using it - as long as the tool does not get very wide-spread, and standard discounts are not based on this detailed stats. If we don't tell what we could, you miss some important information and we feel guilty for that. I believe that in an ideal world CAT discounts should be the basis of negotiation (per project), not the everyday practice. But translation tool vendors hardly have any influence on business practices, I'm afraid.

If you were us, what would you do?

Direct link Reply with quote

Johnny Speiermann
Local time: 13:57
English to Danish
+ ...
No problem in discounted TM rates Jun 27, 2007

I have no problem in offering discounts on repetitions and fuzzy matches when working with translation agencies.

But they only get what they pay for. We often receive projects where the agency claims that we are not paid anything for 100 % matches as the customer is not paying for them.

Well, then they have to take care of 100 % matches by pre-translating the files and making sure that these strings are not opening in the translation tool. Otherwise I will be spending time on these segments, and then they have to pay.

But it all depends on the file preparation done by our customer. We often receive Word documents with text extracted from PDF files. This means a lot of hard line breaks, and then we don't offer any discounts.

I don't agree in having the customer deliver and develop their own CAT tools. This happens, and as we work in software localization we use a lot of different translation tools developed by either the end customer or the translation agency. But these tools are often build by engineers and not with the translator in mind, so they often result in crash bugs that cost us lost time, or they require a lot of mouse work which is not at all optimal.

The major CAT tools do also have bugs, but at least you get to know them over time, and you often find a way to avoid them.

The current hype is to develop web based CAT and localization tools. This could work fine, but even if we have a very fast DSL connection (and we are only two people using it) these tools are often slowing down our work.

Direct link Reply with quote

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

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

Cat tools, globalization and the big money grab

Advanced search

Déjà Vu X3
Try it, Love it

Find out why Déjà Vu is today the most flexible, customizable and user-friendly tool on the market. See the brand new features in action: *Completely redesigned user interface *Live Preview *Inline spell checking *Inline

More info »
SDL Trados Studio 2017 Freelance
The leading translation software used by over 250,000 translators.

SDL Trados Studio 2017 helps translators increase translation productivity whilst ensuring quality. Combining translation memory, terminology management and machine translation in one simple and easy-to-use environment.

More info »

  • All of
  • Term search
  • Jobs
  • Forums