Q&A from Fire Ant & Worker Bee March 21, 2013
Posted by Jill (@bonnjill) in Business practices.trackback
The latest Accurapid Translation Journal has a very interesting Q&A about Trados pricing in its Fire Ant & Worker Bee column (which is always an enjoyable read). Since I have recently received similar requests from agencies (whatever happened to the good ole 30/60/100 pricing scheme Trados used to suggest?), I was very interested in reading the answer and thought you might be as well.
Q:
Dear Fire Ant & Worker Bee,
I have been contacted by three different agencies over the last few weeks proposing the following table for CAT tool discounts, all of whom I have refused while remaining icily polite:
Repetitions @10%
100% Matches @15%
95-99% Matches @20%
85-94% Matches @50%
75-84% and below @66%
50-74% and below @100%
No Match (New Words) @100%
Unedited Text in Graphics @100%
There appears to be some company trying to push this grid along with their CAT tool marketing. It is particularly derisory because low-grade fuzzy matches are in reality practically worthless, often costing more time than they save, especially for those of us who use voice recognition. I often set my CAT software to ignore them.
I am writing in case there is anyone new to the profession who is inclined to believe the sales pitch that this is some kind of “industry standard”. It certainly is not. The supplier of a service sets the price, not the buyer. The buyer decides whether or not to buy.
Puh-leese
A:
Dear No Grid,
We agree not 50 nor 66 but 100%, sir, and applaud your reminder that this grid is a negotiating tool—some might say weapon—and definitely not an industry standard.
Self-assured claims to the contrary come from vendors applying commodity-based business models. They are understandably desperate to lock in margins at the low end of the market, where prices are very definitely under severe pressure.
As you probably know, many skilled professionals insist that translation technology is above all a quality assurance tool for ensuring consistency. As one observer notes, “real-time savings are achieved consistently only with large blocks of 100% context matches.” And in other cases? Well, no one is saying that time might not be saved in some instances, with some texts. But that is not what “industry standard” grids—applied across the board—are.
This may be a good time to point out how much more sense it makes to bill by the hour, which recognizes genuine productivity-driven savings, however achieved.
A top interpreter once told us he developed the concentration he needed to perform at the highest level in the booth through intensive chess competitions. We find ourselves wishing translators would play more poker, to gain practical experience in the skills needed to call a negotiator’s bluff.
FA & WB
Musings from an overworked translator 
I’ve never worked with these systems. Not yet, anyway. What interests me is the point about how “real-time savings are achieved consistently only with large blocks of 100% context matches.”
I guess what I’d like to know is exactly where those savings begin. I mean, roughly how large of a text containing roughly how many such context matches do you need to have before it is worth it to even start using these systems in the first place?
Hi EP, in response to your question, I translated a ppt presentation last week with several grids for a market research company. They conducted a study across countries and the country names alone added to about 2,000 words. It would normally take me about 2 hours to translate that, if it were normal text, but as it involved only repetition, it took me less than 15 min (after I’d translated the first set of country names, it was all repetition). I hope that answers your question. CAT tools do save time when there is a lot of repetition.
Don’t even get me started on so-called repetitions which are entirely context-dependent, and therefore not repetitions at all! Or those lovely 80% matches consisting of three German words where the middle word is “des” or “der”…