Question about reduced rates for fuzzy matches with CAT Tópico cartaz: jmartinalvarez
|
|---|
Hi all,
I've been working as a translator for a few years, but I've only become a freelancer recently. I've decided to jump in with both feet and bought a copy of Trados Studio 2015, since I will most likely start working soon on a project that requires it.
I know that's been a lot of debate on whether or not offering discounts for fuzzy matches - this post is not about that. Rather, there's a small technicality that I still haven't been able to figure out.
... See more Hi all,
I've been working as a translator for a few years, but I've only become a freelancer recently. I've decided to jump in with both feet and bought a copy of Trados Studio 2015, since I will most likely start working soon on a project that requires it.
I know that's been a lot of debate on whether or not offering discounts for fuzzy matches - this post is not about that. Rather, there's a small technicality that I still haven't been able to figure out.
My customer will provide me with a TM (I am aware that I should have a look at the TM before starting the project, although I'm pretty sure it's reliable). The customer has also requested a Trados grid with my rates for fuzzy matches in all bands, which I've been happy to provide.
Here's the source of my confusion. I will obviously be using a different rate for, say, segments that have a 75%-84% match *in the provided TM*. At the same time, I will be creating a new TM for this particular project, so I will be using both (the provided one for reference and the new one to, well, help me translate). What I'm not sure about is whether that 75%-84% rate should also apply to segments *inside* the project, which will appear in my newly-created TM. So, for example, if I translate a segment and I get an 80% match for that segment a few segment below, I'm not sure whether the 75%-84% rate should apply to that translation.
My gut says no - the customer is just expecting a slight discount for texts that they've already provided me with, and the analysis will be made at the beginning of the project, comparing the text against the provided TM. However, since I've been out of the freelancing world for so long I'm not really sure what the conventions are, or whether there is a standard at all, so I'm sorry if the answer is obvious.
Any help is very much appreciated! ▲ Collapse
| | | | Christine Andersen 
Dinamarca
Local time: 00:35
Membro (2003)
dinamarquês para inglês
+ ...
| Go with your gut feeling | Jun 29, 2016 |
In the end, you are the one doing the work, and you should be setting the rates.
Ultimately, you want a viable rate for the whole job, and how you reach it is less important. Whether you give a further discount for these 'internal repetitions' or not is up to you.
How much work is involved: are they replicated automatically without effort, like 'Page 3/5 etc.? If they are things like headers and footers on pages, which may or may not be counted, depending on how the te... See more In the end, you are the one doing the work, and you should be setting the rates.
Ultimately, you want a viable rate for the whole job, and how you reach it is less important. Whether you give a further discount for these 'internal repetitions' or not is up to you.
How much work is involved: are they replicated automatically without effort, like 'Page 3/5 etc.? If they are things like headers and footers on pages, which may or may not be counted, depending on how the text is formatted among other things, do you have to check for variations like page numbers and minor variations, or are they all completely identical - just a logo and file name, for instance?
Consider things like that and how much work is involved in 100% matches and fuzzy matches.
In my language pairs fuzzy matches are quite useful, because there are very few inflections, but in some languages they are practically useless, because you end up re-typing the whole segment anyway.
Decide what works best for you, and set your base rate accordingly. ▲ Collapse
| | | | | I'd say that includes everything, "old" and "new" TM | Jun 29, 2016 |
jmartinalvarez wrote:
My customer will provide me with a TM (I am aware that I should have a look at the TM before starting the project, although I'm pretty sure it's reliable). The customer has also requested a Trados grid with my rates for fuzzy matches in all bands, which I've been happy to provide.
Here's the source of my confusion. I will obviously be using a different rate for, say, segments that have a 75%-84% match *in the provided TM*. At the same time, I will be creating a new TM for this particular project, so I will be using both (the provided one for reference and the new one to, well, help me translate). What I'm not sure about is whether that 75%-84% rate should also apply to segments *inside* the project, which will appear in my newly-created TM. So, for example, if I translate a segment and I get an 80% match for that segment a few segment below, I'm not sure whether the 75%-84% rate should apply to that translation.
Just to make an example, most word-counts nowadays also include an "internal" matches analysis, that is to say, the number of repetitions etc. within the given text (therefore, even without a TM). Anyway, more in general, if you're willing to give discounts based on previous translations whose quality/reliability you're not even sure of, then all the more reason to apply the same discounts on matches/repetitions based on your own translations (whose quality you're sure of...), don't you think?
At any rate, if there will only be a single global analysis at the beginning of the project, as you were saying (vs. multiple analyses for various batches as the project progresses), then this should not be a problem for you in this specific case...
| | | | | I'd say things have to be balanced | Jun 29, 2016 |
We freelancers have so many expenses.
I for instance pay each year: 70 € for web page and professional e-mail hosting, 125 € for maintaining my CAT-tool updated and with technical support, 100-200 € for specialised books and webinars, 100 € for my membership in the Spanish National Association of Translators, Interpreters and Proofreaders, the fees of the economist who presents my declarations before the Tax Authorities, I have to pay for my own vacations (not like an emplo... See more We freelancers have so many expenses.
I for instance pay each year: 70 € for web page and professional e-mail hosting, 125 € for maintaining my CAT-tool updated and with technical support, 100-200 € for specialised books and webinars, 100 € for my membership in the Spanish National Association of Translators, Interpreters and Proofreaders, the fees of the economist who presents my declarations before the Tax Authorities, I have to pay for my own vacations (not like an employee), and I need my vacations because I have two small children, I have to pay all my taxes and social security fees on time, and I still was not able to get the professional liability insurance I would need, apart from all this, there is no guarantee that next month my income will be the same as this month... and the list really goes on and on...
For all the above described I offer very few discounts, for 100 % reps (only against MY OWN previous TM + internal 100 % matches against MY TM); depending on the client I might offer the translation of +95 %-matches for free, because there will be a lot of 93 %-matches that I will charge at my full rate, and in these cases this really is the correct thing to do from my point of view. But these are rare occasions, for example in the case of texts that are quite repetitive (data protection regulations, etc.).
Generally I would say offer only the minimum discounts, because we have to cover so many expenses and somewhere we have to gain that money to be able to pay for everything and make a living as well. It is like a balance – on one side you gain (your CAT-tool helps you work more efficiently), on the other you lose (you have many expenses as a freelancer which you don’t have as an employee).
Hope this point of view helps you to take your decision about how to achieve the ideal balance in these “freelancer-issues”.
Best regards,
Agnes
[Edited at 2016-06-29 13:38 GMT] ▲ Collapse
| | |
|
|
|
Hi,
Thanks a lot for the replies, they help a lot! I see things more clearly now.
| | | | | | Jo Macdonald
Espanha
Local time: 00:35
italiano para inglês
+ ...
| Cat discounts on client's TM matches only | Jun 29, 2016 |
jmartinalvarez wrote:
Here's the source of my confusion. I will obviously be using a different rate for, say, segments that have a 75%-84% match *in the provided TM*. At the same time, I will be creating a new TM for this particular project, so I will be using both (the provided one for reference and the new one to, well, help me translate). What I'm not sure about is whether that 75%-84% rate should also apply to segments *inside* the project, which will appear in my newly-created TM. So, for example, if I translate a segment and I get an 80% match for that segment a few segment below, I'm not sure whether the 75%-84% rate should apply to that translation.
When quoting for a job I'll apply cat discounts to matches from a TM the client can provide. If I can get more matches from my TM that's my business. Also any leverage I can use from segments I've already translated while doing the translation, again that's my business.
Watch out for some clients who will create an empty TM and run the file/s through it to get "internal fuzzies" which are basically similar segments to ones you will eventually translate. Imo these are No matches until I've translated them so any "internal fuzzies" I'll consider No matches = 100% of my rate.
My CAT rates
(Percentage of standard rate)
Exact match (100%) + Repetitions -> 30%
Fuzzy match (75-99%) -> 60%
No Match (0%-74%) -> 100 %
| | | | | sharing the benefits of progress | Jun 29, 2016 |
I think this is called 'homogeneity' in memoQ, no idea about Trados though.
I include this myself, since fuzzies within a project save me time and work.
It's the same if I have 4 tyres changed down at the garage, I don't expect to pay 4 times the price of a single tyre. It's only fair to share efficiency savings, that's how the world progresses. Sadly, I expect many here will disagree and shoot this down!
BTW fuzzies below 75% are generally useless, and wi... See more I think this is called 'homogeneity' in memoQ, no idea about Trados though.
I include this myself, since fuzzies within a project save me time and work.
It's the same if I have 4 tyres changed down at the garage, I don't expect to pay 4 times the price of a single tyre. It's only fair to share efficiency savings, that's how the world progresses. Sadly, I expect many here will disagree and shoot this down!
BTW fuzzies below 75% are generally useless, and will save you little if any time. ▲ Collapse
| | |
|
|
|
| A rate can imply all that | Jun 29, 2016 |
Hello, just like some clients want to be invoiced tax included, you can provide a rate per word that is inclusive of match/no match discount or simply mention a NET rate and specify that you do not charge for the TM provided through your work. The text that you do not get paid for, you still have to type/accept/read or change it.
| | | |
Richard Purdom wrote:
... Sadly, I expect many here will disagree and shoot this down!
Sharing the benefits of progress when sharing the expense of software and updates.
Philippe
| | | | | Degree of usability of fuzzy matches | Jun 29, 2016 |
Richard says that fuzzies below 75 % are generally useless. May I add that I think this depends on the type of text AND the languages one works in.
In the case of the one I quoted above (data protection regulations), this may be true to some extent, but I can assure you that in 90 % of the highly specialised German legal texts that I translate into Spanish, a fuzzy match of 85-95 % is most of the time really useless, as I have to make sure that the legal value/effect of each expre... See more Richard says that fuzzies below 75 % are generally useless. May I add that I think this depends on the type of text AND the languages one works in.
In the case of the one I quoted above (data protection regulations), this may be true to some extent, but I can assure you that in 90 % of the highly specialised German legal texts that I translate into Spanish, a fuzzy match of 85-95 % is most of the time really useless, as I have to make sure that the legal value/effect of each expression/term/word/sense is correctly transmitted and I gain almost nothing with the fuzzy match (in terms of time used for translation).
And what we are paid for is our knowledge/skill which we employ spending our working time, in the end I think we are paid for our knowledge + our time, for both things. And if I gain no time with the fuzzy match, I cannot charge less, not because I do not want to be flexible, but because of what I explained before.
So it really depends on the text and the language, I think.
[Edited at 2016-06-29 15:14 GMT] ▲ Collapse
| | | | Samuel Murray
Holanda
Local time: 00:35
Membro (2006)
inglês para africâner
+ ...
jmartinalvarez wrote:
I will obviously be using a different rate for, say, segments that have a 75%-84% match *in the provided TM*. ... What I'm not sure about is whether that 75%-84% rate should also apply to segments *inside* the project, which will appear in my newly-created TM.
Although internal fuzzy matching has become a feature in some CAT tools in in recent years, I'd still say that unless the client specifically says that internal fuzzy matches are discounted, then they aren't. In the old days, only external fuzzy matches (i.e. from the client's TM) were discounted -- not for logical reasons but simply because CAT tools in the old days did not really count these things.
My opinion on this hasn't changed (also noted in some of the threads that Philippe mentioned), namely: if the reasoning for fuzzy match discounts is to charge less for segments that take less time to translate, then you should be charging a higher rate (i.e. give less discount) to fuzzy matches from an external TM than from your own TM, because you can "trust" your own TM and you know your own terminological preferences already, whereas a client's TM often has questionable pedigree, and regularly requires more careful examination and/or editing for the same match percentage.
That said, at least one of my clients expects me to give discounts for segments that come from the TMs that I had generated from their previous translations for the same end-client, but does not expect discounts for segments from within the current translation. In other words, only external matches are discounted, but my own old TM is used as the "external" TM. It saves the client having to send me the updated TM every time, since my own TM is the "updated TM" anyway.
| | |
|
|
|
Mervyn Henderson (X)
Espanha
Local time: 00:35
espanhol para inglês
+ ...
| As Philippe says ... | Jul 1, 2016 |
... the reason a discount is given is because you are supposed to work FASTER, otherwise what's the point? I recently questioned a company's internal CAT, a real clunker not much different from Trados Workbench, which has so many hoops and slowdowns in it that, as often as not, I spend MORE time on the ruddy job. Negotiations pending to use a reasonable CAT.
Also, how do you know what shape someone else's TM is in? Is it full of Google Translate? Is it full of source segments simply... See more ... the reason a discount is given is because you are supposed to work FASTER, otherwise what's the point? I recently questioned a company's internal CAT, a real clunker not much different from Trados Workbench, which has so many hoops and slowdowns in it that, as often as not, I spend MORE time on the ruddy job. Negotiations pending to use a reasonable CAT.
Also, how do you know what shape someone else's TM is in? Is it full of Google Translate? Is it full of source segments simply copied but not translated to target, but marked up as 100% translations? Has it been mauled by many mitts over the years with no standardisation of the terminology used? On many occasions I've had to demand full rate because of all the weird stuff I've had to reshuffle.
[Edited at 2016-07-01 08:19 GMT]
[Edited at 2016-07-01 08:32 GMT] ▲ Collapse
| | | | | Thanks for the feedback | Jul 1, 2016 |
Lots of good replies here!
In the end I clarified this directly with the customer, but I'm glad to see that my confusion was at least partly justified. I will be charging 100% for internal matches and provide a modest discount on fuzzy external matches, of which I actually don't expect many because of the characteristics of the project.
Thanks everyone for your feedback, it's been really helpful!
| | | | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » Question about reduced rates for fuzzy matches with CAT | Anycount & Translation Office 3000 | Translation Office 3000
Translation Office 3000 is an advanced accounting tool for freelance translators and small agencies. TO3000 easily and seamlessly integrates with the business life of professional freelance translators.
More info » |
| | TM-Town | Manage your TMs and Terms ... and boost your translation business
Are you ready for something fresh in the industry? TM-Town is a unique new site for you -- the freelance translator -- to store, manage and share translation memories (TMs) and glossaries...and potentially meet new clients on the basis of your prior work.
More info » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |
Login to reply/comment 
Translation news
