Sep 28, 2009

Four months of Twitter

Last May I finally decided to become a tweeter on Twitter (or a twit some might contend). Four months on, my use and perceptions of that medium have changed little. I follow few and am overwhelmed by those feeds , so I simply can't imagine how anyone following hundreds can pan the gold out of all the mud. Major filtering tools I suppose. However, I am finding it an interesting channel of communication, particularly with colleagues and strangers who go light on the self-promotion and details of when they take showers and point me to interesting resources. I also appreciate the pointers to news I might otherwise miss.

It's a medium that is clearly abused by some. I killed one feed, because the woman writing it seemed to be intent on retyping an entire book of crappy aphorisms. Multilingual feeds in languages that I don't understand are occasionally frustrating (and I'm sure some reading my feed feel that way about my German tweets), but it's also fun to puzzle out the odd tidbit of Portuguese or Italian with no pressure. Lately I've been getting spammed by "nice" girls offering themselves for sale, and I wish I knew a way to block nonsense like that and not restrict others with reasonable intentions.

On the whole I think that Twitter has been a worthwhile addition to my communication toolset. I'm not sure where else I could get a bunch of good, fresh tips on children's book reviews and publishing like Carsten Peters (calutateo) provides or follow the tasteful reading lists of others. Since I'm not selling anything at this point (plans for a technical book have been delayed indefinitely due to surprise developments in the technology in the past year), I can't confirm or refute any of the claims of Twitter's importance for marketing, and I really don't care about that point. I'd rather trade work philosophy one-liners with colleagues, point to useful sources for self-education and have a bit of fun.

Sep 24, 2009

Toss a grenade in the network

Over the course of years, errors have accumulated in the Windows registries of our working computers. So we were informed by RegistryBooster, a tool from Uniblue, which I tested as part of a WinZip promotion. I have a natural suspicion regarding utility software, which comes from nearly over three decades of seeing the damage it can do if there are differences in one's system compared to the systems for which the tool was developed or on which it was tested. Today's nifty partitioning tool might be tomorrow's data killer. And given the value of my data archives, this inspires a lot of caution.

However, sometimes caution gives way to frustration. Although little software has been added to our systems since they were initially configured, enormous amounts of disk space has mysteriously been gobbled up by some process, and system responses have gone from snappy to very, very sluggish. Registry problems seemed a plausible explanation for at least some of this. So we bought a license for Uniblue's tool and tested it on one of the production systems, unfortunately at a time when a lot of work was scheduled. (Yes, I know this isn't a good idea. I found out about the action while it was in progress.) The initial results looked good - performance did in fact improve a bit. However, suddenly the shared folders on the "repaired" system were no longer accessible.

Since I no longer deal with niggly system software issues on a daily basis, it took a while before I figured out that the helpful system utilities from Uniblue had trashed the name of the workgroup on our network and activated the computer's firewall - without asking I am told. Not nice, really. For the average user without a lot of network expertise - and I'm not far from that category these days - this is a bit like tossing a grenade into the network. Certainly it bombed my productivity for a long afternoon.

The real moral of the story, though, is never, never, never make major interventions in your computer system software when deadlines are looming!

Sep 23, 2009

Infix PDF Editor: useful for some jobs

A few years ago a Brazilian colleague of mine recommended the Infix PDF Editor from Iceni Technology with great enthusiasm. I tried it at the time, but it was obvious that we had very different PDF types that we worked with and different philosophies of translation, so I concluded that the tool was of little value to the translator except in special situations, and I set it aside. What were those situations? For me, it seemed a decent tool for minor touch-ups of a PDF about to go to print (but I don't do a lot of pre-press proofreading), and for translating simple flyers where work with a TEnT makes little sense it also seems useful. For larger documents I did not see the value, because typeovers of large blocks of text in a PDF document, where possible, run just as contrary to my methods of working as typing over large chunks of text in MS Word or another environment. There is too much risk of content being skipped or deleted, and there is no access to integrated TM and terminology tools.

Recently, however, I discovered another area in which PDF Infix Editor makes a very useful addition to my toolbox. Occasionally I am asked to translate large batches of engineering drawings which have been scanned and reduced to A4 in PDF files. These do not contain editable text. The quality is also usually so miserable that one can forget OCR. Years ago I developed a procedure which the client is fond of: I save the individual pages out as graphics, embed them in an MS Word document and then overlay the portions to translate with text boxes in Word (with appropriate opacity settings and rotation). This can be quite time-consuming.

When another batch of some 120+ of these awful documents arrived recently, I thought of trying the Infix PDF Editor, because I remembered that it had a decent text tool. By overlaying white rectangles followed by text boxes, I was able to accomplish the translation task in the PDF document with less effort than required for my previous work in MS Word. Here's an example of the change:

This saves me time and irritation, and the client will save money. Everybody wins.

Since my initial brush-off of the tool I have also discovered a few other positive aspects for translators that overcome some of my initial objections: text copied and pasted from the Infix interface to MS Word or another environment does not contain all the awful line breaks that one usually sees when copying from Acrobat reader. And text pasted into the Infix interface takes on the properties of the text segment over which it is pasted. This has the advantage that I can copy awful, small text out of a PDF in Infix, paste it into a DOC or RTF file, change the font size to something my old eyes can read and translate it there or in another environment such as DVX or MemoQ (where the font change is unnecessary). Then I can paste chunks back into the original PDF and the fonts and formatting will be largely correct. Mind you, this can be an awful way to work in a large document, but it has advantages in some situations. And I can, in fact, by copying out and pasting in, make use of by favored TEnT methods, with the benefit of translation memory and terminology management.

So now I would say that the Infix PDF Editor is in fact a useful addition to the toolbox for handling PDF, which still must include a first-class OCR tool, such as ABBYY FineReader or OmniPage. PDF is by no means a uniform format but rather a "wrapper" for making many formats accessible to readers, and it will continue to be a time-consuming challenge to translators which must be reflected in appropriate service charges for the extra effort involved.

Sep 22, 2009

Planning to fail

There are days when I think Armageddon would be a nice alternative to yet another urgent emergency translation delivery. There are days when sleeping under a local bridge, barbecuing sausages over an open campfire and listening to the rustling of wild boar in the bushes seem a desirable change from an office piled high with dictionaries, technology new and old and - my favorite - ringing phones. Four of them. Five if you count the fax. There are days when I hear the words "surely, you can fit this one in somewhere" and know just exactly where I would like it to fit.

There are days when failure seems like a pretty cool option.

For those of us tired of freelance success, there is help to be had. A number of colleagues offer advice on how to extricate oneself from the trap of success. A recent contribution is the excellent self-help article 10 Ways to Make Your Freelance Business Fail, though classics like 40 Fabulous Faults of Freelance Failures are old sources of inspiration not to be overlooked. These can be useful references even if you haven't decided yet that life in the soup kitchen line would be a lot less stressful.

One of my favorite lines of inspiration from the second source is something to remember when you consider continuing education options:
Why should you get certified when you’re already certifiable?
Indeed. A point worth considering carefully.

When I follow many of the online discussions on ProZ and elsewhere, it's clear to me that many colleagues and wannabe colleagues have discovered such useful advice long ahead of me and use it as the "secret weapon" pointed at their own temples with the thought of putting themselves out of the misery of being independent business people responsible for their own fates. It really is much better to live in a tightly controlled environment with strict rules and assigned roles, one in which only electricians may screw in light bulbs and only those with degrees in translation are permitted to translate complex procedures involving the handling of toxic chemicals or radioactive materials. Prices should be set by a Higher Power (that's HP, not HD) and apply universally - in every country - so that everything is fair. (Perhaps some mechanism will have to be found to keep those living in the "developing world" from living above their station, but the main thing is that we concern ourselves with these important, fundamental principles and shift the focus away from mundane matters like reality and how to compete effectively in a modern economy.)

Unfortunately for my long-suffering clientele perhaps, I usually awaken from these Walter Mitty as hobo reveries and realize that I actually like what I do and the people I do it for and just need a bit more time off occasionally. So if I start blogging from the Bahamas for a year, you'll know why.

Sep 18, 2009

In search of the perfect caffèlatte

There are rumors in Budapest that this blog is powered by Unicum. That is only occasionally true. For the most part, inspiration comes from steady, massive doses of caffeine, administered in a variety of ways as teas or coffee, most especially caffèlatte. If I'm feeling ambitious I might do a nice cezve of Turkish coffee, but my comfort drink is the latte, with strong coffee made in my French press. My espresso machine and steamer died long ago (worn out), and I prefer the greater control that the press gives me over the quality of the brew. For quite a while after the steamer got tossed out I used an electric whisk to foam the milk to varying degrees of stiffness until I discovered the pleasures of milk foam prepared in a ceramic frother for hot chocolate that I bought under the mistaken impression that it was another version of a French press. There are apparently milk frothers widely available for this very purpose, and foam prepared in this manner has superior stability and texture. Foamed hot chocolate milk in combination with coffee so strong you can stand a spoon in it is definitely the thing to keep one charged up for a long evening of patent translation. And usually the preparations are topped with a grind of spices obtained from our local Viennese coffee house.

I find that the rituals in our office for preparing teas and coffees, with each of us relieving the other under pressure with a fresh cup of a favorite brew, provide a bit of needed sanity and friendly communication on days when it's hard to speak two sentences before the phone rings.

Rates and guidelines for writing and translation

Translators are often obsessed with size, in this case the size of the fee that can be charged for work. It's understandable: we all have to eat and most of us have to pay for at least part of the food and perhaps the roof over our heads. And insurance, software, equipment, electricity, water and more things than I care to remember. For reasons I fail to grasp completely, many translators have a hard time figuring out what to charge for their work and look for comparative data, then argue about whether it is even relevant. When I talk about rates in Germany as reflected in the BDÜ Honorarspiegel, some scoff that these are obvious lies, because nobody can get that much in today's global economy, while others (perhaps financial translators) snicker at the thought of qualified colleagues prostituting themselves for pennies (albeit a lot more pennies than you'll see offered in a typical ProZ job post).

This afternoon I followed a link from a discussion thread to an interesting page from the Society of Authors in the UK, which offers some guidance on pricing for translations of various kinds. From there I ended up somehow on another page with suggestions for minimum freelance translation rates for various language categories. So the next time, dear Reader, when you are tempted to take that fantastic opportunity from a UK agency for twenty quid per thousand words, have a look at these pages (which do not show particularly high rates) and think again. Oh yes... hourly charges are discussed as well, albeit rather low ones. For those who don't think in GBP, have a look at and convert the numbers to your preferred currency.

Language Service Provider Network

A week ago at the ProZ powwow in Berlin I had a chat with financial translation specialist and outsourcer Ralf Lemster, a colleague known for his integrity and no-nonsense business acumen. He knew that I have been searching for some time for a good solution to our internal administration problems, which aside from my preference for translation work and copywriting over writing invoices arises to a large extent from the fact that I have yet to discover a networkable solution for project management and billing that is affordable and would work well for a partnership of two freelancers with an assistant to sort paperwork and wash windows so we can see that there is indeed a world outside our office apartment. For years the software we used has resided on my hard drive and is not accessible to my partner. (We tried putting it on a third machine, but this had severe drawbacks too.) This has led to "disconnects" in communication and tracking of jobs, and the processes we have used involve unnecessary steps that become quite burdensome with the rapid pace at which this office frequently runs. Even my sanity-saving folder organization system breaks down sometimes, because inquiries come in too quickly to keep things sorted before the next query comes.

Ralf mentioned that a partner company of his in Berlin is about to release an excellent solution for online management of a freelance translation business or small LSP. He has invested a lot of time in his own management systems, and when he mentioned the intention to switch to this provider my attention was fully focused by this surprising comment. Some people will chase after any chimera or change systems like Twiggy once changed clothes. (Sorry, I don't follow fashion, so I have no idea who the top models are these days.) I don't think he's one of these people. So I gave the company a call and had a long, friendly chat with the CEO, who described a system which sounds like my dream for the past few years. It's scheduled for release in January, but there is a pilot phase planned to start in October, and it may be possible for interested parties to participate and "kick the tires" for free. After that the monthly fees will be something like € 29 per login, much less than for other online solutions like Worx or BeeFlow. So I signed up for the pilot at, and as of October (or whenever things get off the ground) I'll try to run our business using it and report here on the triumphs and tribulations. I would encourage others with similar needs to have a look at the web site, learn about the available features of the system and give it a try if it sounds interesting.

Sep 16, 2009

Curiouser and curiouser

The ProZ censors are at it again, and to be quite honest, I can't see what for. Today I received the following message for a post made some time ago, which was purged today:

Dear Kevin Lossner,

This message is to inform you that your post "Really?" has been removed from public view because it was not in line with site rule

Site users are expected to treat each other with courtesy, whether posting publicly or making direct contact.

Thank you in advance for your understanding and future cooperation.

Jared moderator.

Since there was no link given I had no idea what the allegedly objectionable content was, so I filed a support request to ask. Jared kindly responded and copied the text out along with the following message:

Text of removed post
Dear Kevin,

I'll copy the text of the post below. Note that, in the case of removed posts, no links are sent since the posts are no longer visible.


Samuel Murray wrote:
Exactly. That is why the system isn't a quoting system -- it's a bidding system. But someone at thought that "bid" had a bad ring to it, so they called it a quoting system.
Is this firm knowledge or speculation? Doesn't sound bad to me either way. In an RFP, which in a sense is what many of the job posts could be considered, the terms "proposal", "bid" and "quote/"quotation" are really equivalent, and I think my friends on the big island off the coast of France are fond of "tender" as well. (That might come from all the tenderizing that is said to have gone on in the public schools in days presumably past.)

As you know, some people claim that these job posts are like slavery and ought to be banned from ProZ. I vehemently disagree. I think we should encourage slavery in the profession and therefore I herewith propose to Henry and the staff of ProZ that programming resources be dedicated to creating slave auctions. We won't need dental X-rays and beefcake photos of those up for sale - a summary of their qualifications and a commitment for a unit of capacity within a specific timeframe is all that would be needed. (Sorry, this is only part-time slavery... those who want it full time will have to look for a day job.) Outsourcers could then bid - quite publicly - for the advertised capacity for a specific translator. Those on the block could even specify the starting bid level to make it "fair" :-) Whatchy'all think?


Member services and support

The "rule" allegedly violated reads as follows:
Mutual respect, professionalism and fair play are expected. Site users are expected to treat each other with courtesy, whether posting publicly or making direct contact, and are advised to act under the assumption of good faith. Harassment of, or attacks on, individuals or groups, of any form, as well as discouragement of another's use of the site, will not be tolerated. No action aimed at gaining unfair advantage in KudoZ, the directory or elsewhere, whether taken alone or as a group, will be tolerated.
Perhaps I'm merely getting old and senile, but I fail to see the connection here. But as Barney Frank would say, trying to have a conversation with some people is like trying to talk to a dining room table. Jared is a nice fellow, but I think he's off in the Twilight Zone here with his RuleZ.

I was actually serious with the slave auction proposal in any case. I think it would be fun, and it would be revealing in a lot of ways. I remember years ago too that this used to be a popular fund-raising method for churches and youth clubs in the California town where I grew up. The other good thing about this proposal is that the bidding structure would move prices up and reveal to some of the bottom feeders what others are willing to pay. Instead we have the usual system where hungry translators in some language pairs compete to see who can work for the fewest calories per day. Better to be open about that slavery thing, because real slave labor has more value than what many ProZ outsourcers are willing to pay.

Sep 15, 2009

What DVX users are saying about MemoQ

Some snippets from today's Yahoogroups digest for the dejavu-l list:
I've just added my name to the group buy too - I don't much like the idea of having to get to grips with another CAT, but do want to encourage developers who listen to translators' needs, and everybody seems quite satisfied with MemoQ for the time being. I do love DVX (compared with Trados), but I do run into a few bugs occasionally and have also been waiting for their new release. I tend to share other users' reservations regarding Powerling...

Same here.

As Frank mentioned the possibility, I contacted Proz to inquire about the TGB for MemoQ and got the following info:

I have extended the time to add your name on the TGB list.
Please feel free to add your name here on this memoQ TGB.

The sale will close on 21 September or when 40 buyers are added to the list.

And so I added my name as no. 29 on the list.
And then less than an hour later...
There are now 32 people signed up, so maybe we will get the 8 more needed.
I wonder if Atril is paying any attention to all these DV users wanting to buy MemoQ.
Five minutes later:
make that 33.
One DVX user was desperate enough to get MemoQ with active development and real support that he commented shortly afterward with regard to the possibility of the magic number of 40 orders not being achieved for the ProZ group purchase:
If not, I'm prepared to buy two.
Over the last few days, I've been shocked to see how much of the message content on the Déjà Vu user support forum has focused on the desirability of purchasing MemoQ, because Atril is not moving at the pace required to meet the needs of its very loyal user base. I too am an enthusiastic user of Déjà Vu software from Atril, and I can bore people all evening with stories of how that software has helped my business over the past decade, how the company founder once saved my bacon at 3 am by fixing a corrupted file for me and so on. Most long-term DV users have stories like mine. The problem is that we are speaking of the olden, golden days. Ain't much been happenin' these past five years or so. Waiting for a new DVX release seems almost more absurd than waiting for Godot. It shouldn't be this way.

There was much fanfare about the new marketing and support relationship with PowerLing, and after I was publicly critical of the lack of anything apparently useful coming of it, I received a nice phone call from that Atril partner asking me to give it all a chance and be "patient". Well, all DVX users have a lot of practice with patience in recent years. The real problem, however, is not PowerLing's competence or lack thereof, it's the dearth of communication and lack of visible progress from Atril itself. A company has to work hard to destroy the trust of a loyal user base like Atril has (had?). Imagine the hordes of Apple fanatics suddenly finding religion with Microsoft or some geeky Linux environment. Imagine Toyota owners lining up to buy GM cars. Not even in an alternate universe. But this seems to be the direction that things are headed with Atril if the long-overdue releases don't start happening. Go ahead, charge us upgrade fees. Please. Just release the damned software at last so we have to be less creative with workarounds as the release versions get left farther behind in the technology race.

MemoQ 3.6 now offers the equivalent of DV's autoassemble feature. External views with exportable comments are coming soon. Inclusion of unsegmented content from Trados jobs (so dates and numbers can be handled) has been around for a while, the same for the sophisticated TM-driven segmentation technology that enables better leverage that you'll see from any other CAT tool on the market that I am aware of. DVX still has a few advantages, but with the Hun(garians) at the gate and the walls crumbling, it's time to mount a real defense of market share, not some silly "I Love" slogans like PowerLing has offered. Real technology. Real progress. Let's see a healthy competition between two great companies who can cheerfully toast each other as they carve fat chunks out of the SDL's market share.

This evening there are 38 of 40 units claimed in the MemoQ group buy at ProZ. Most of the final push (maybe all of it) is from DVX users. Atril, please wake up.

Sep 14, 2009

My weekend on the fringes of the BDÜ conference

This weekend was quite a fine one for meeting old friends and acquaintances and meeting other friends I know through online media in person at last. The BDÜ conference Übersetzen in die Zukunft - Herausforderungen der Globalisierung für Dolmetscher und Übersetzer drew over 1600 attendees for three days of presentations, exhibits and events. As my friend Aniello Scognamiglio related last night while we enjoyed a quiet meal at Wittenberger Platz, the conference was a huge success. Hearing this from someone who is sometimes skeptical of the BDÜ and having heard this from equally discerning colleagues Friday night at the ProZ PowWow at the restaurant Luise near the FU Berlin I can only conclude that it was something very special, which I hope will be repeated in the future. I had planned to attend myself, but the hunting course I started in August conflicted with two of the scheduled conference days, so that was not possible.Too bad, because as the event was described to me, it sounds like the BDÜ got everything right.
BDÜ Conference in Berlin.
Photo copyright (c) 2009 Aniello Scognamiglio

The ProZ gathering Friday evening was smaller than the rather large get-together that Steffen Walter has organized for years, but it was still well-attended to the point where in a long evening I didn't have time to talk to half the people I wanted to. Say what you like about the crappy job posts from bottomfeeding Chinese and UK agencies on ProZ, the juvenile behavior of many unqualified moderators and the often almost abusive lack of transparency in many matters, the powwows organized by members are inevitably worthwhile. I always meet interesting, competent colleagues and learn something from our exchanges. This time around I had the pleasure of a long chat with Oleg Rudavin, author of an excellent book on Internet freelancing for translators (which I'll be doing a full review of for the BDÜ sometime in the next few months). He was at the conference to promote his book and represent ProZ, and our conversation made it clear to me that the insights in his book are drawn from a deep well of varied and interesting experience from which there is much more to be had. He is considering German and French versions of the book adapted to the respective markets, and I do hope these plans are realized, because they will - at least for German - fill a significant information gap for translators without strong English skills. I also learned about a good possibility for online business mnagement for freelancers (more on this later maybe after I check out the details), and my dog Ajax and I very much enjoyed the company of agency owner Caroline Montexier from CMLT in Cologne who updated me on the market situation for agencies while Ajax tried to climb in her lap and eat her ponytail. Ralf Lemster had a cool new Australian hat which reminded me hat I need something better to keep the rain out of my eyes while I sit up late at night waiting for the wild boar to show their ugly snouts. I can recommend these gatherings even to dyed-in-the-wool ProZ-haters, because they really are fun, and I always find the opportunity to talk face-to-face with others in my chosen profession worthwhile.

Other than that it's been a quiet weekend at home with my German Wire-haired Pointer Ajax, translating and making applesauce from all the myriad apples falling from the trees in the garden. The other two dogs (wire-haired Magyar Vizslas) have been off in the Netherlands adding to the increasingly crowded trophy shelf in my partner's office and giving me the peace and quiet I need to perfect Ajax's retrieve.

Sep 6, 2009

How low can you go?

Recently a colleague posted the following in a ProZ discussion thread:
... I've been wondering how accurately the "aggregate rate information" at ProZ represents real rates... These are target word rates, but I wonder how much lower actual rates are? I know that I sometimes charge less than my target rate (up to 20% or so) if other conditions of a given assignment make it a good deal overall.
Here we go again. I find the assumption that actual rates are lower to be quite an interesting one, reflecting an interesting mindset. Why should they in fact be lower? "Actual rates" to me reflect those to be had in various market segments, and these tend to be a lot higher than many translators on ProZ assume. Not always, but often enough that I would not make the assumption that actual rates received are lower than the target rate. I myself have a target rate, but given the individual situation of a business transaction I often charge quite a bit above the "target". For me the target reflects an average expectation for an average situation. Just as something may prove simpler and justify a rate cut (like a recent job I did with 53% repetition that I discounted when I wrote the invoice, although my client expected no such break), special complexity (really crappy source scans, difficult texts, extreme amounts of source text errors, dealing with almost anything Swiss, etc.) more than justifies a premium.

The assumption of lower rates isn't entirely daft, however, depending on what exactly you consider the rate to be. With Client A I might charge in full for all source text regardless of any repetitions or TM matches. My effective rate is the same as the simple published rate. For Client B I might have negotiated a reduction in fees for TM matches above 90% if the older material was written by me. In that case, my effective rate is lower, of course, though my earnings for actual effort are most likely not any less. But that is not the writer's point as I understood it.

The assumption that translators will be paid less or should accept less because some sectors of the economy are experiencing difficulty is unfortunate and misguided. Certainly some in our profession are affected, but the challenges faced are very much related to individual situations, business forms and language pairs. Some months ago I had an interesting conversation with an agency owner whom I respect very much, because she has a well-run operation with an emphasis on quality. She complained of a 40% drop in end customer volume, though she was able (at that point) to maintain the rates themselves at previous levels. The primary declines were said to be in minor languages: instead of translating dog food advertising into 15 languages, a client of hers might choose to focus only on 5 target languages and let the Croatians struggle to understand the list of ingredients in Italian or English. Some clients looked for cheaper service elsewhere, but this was less of an issue, because most were too smart to have important marketing materials or technical instructions entrusted to unserious non-professionals willing to take on real jobs for pocket change. Since then a number of my other agency clients have mentioned declines in overall volume, some greater, some less.

Does this affect me? Not really. As another blogging translator (can't remember which - sorry) pointed out recently, agencies have other options to stay afloat, like cutting staff, changing phone plans or moving to less expensive office space. Individual translators have fewer options. The space under my local bridge is already taken up by other hobos, so I prefer to stay in my house. And WLAN in front of a campfire sucks.

But the good translators I know in my language pair are all rather busy. At least they are when I try to refer clients to them, which happens several times a week at least. They might be less busy with particular agencies, but overall they are doing well, and the trend of companies to cut out the middlemen and go directly to a good freelance translator is not hurting them at all. These end customers are more work sometimes, but they pay much better. How much better? One company whose materials I used to translate through an agency approached me a few years after I ended my relationship with the agency because I was deemed "too expensive". I get 58.8% more per line from the direct relationship as well as very generous compensation for revisions, terminology work and other incidentals, and the company is paying a good bit less than the agency charged previously. The project management by the company's employees is better too - questions get answered more quickly without an agency PM sitting on them like a hen on a nest full of eggs.

Better rates like that are actual rates that are part of actual markets. Just like a better rate I might give to a friend who needs a web page translated to sell his puppies or a break I might give to a recent college grad with no job prospects in Germany who has chosen to emigrate to Canada with the clothes on his back. Rather than worry about market generalizations, I think it's more important to look at our own range of services, our personal niche markets (if any), the types of clients we deal with (the BDÜ has 5 categories in its rate survey, but one could just as well define 15 types or more) and research specifics for those markets that we intend to be part of. Anything else is really mental masturbation.

As an example, a young wannabe colleague (recent university grad) posted a note in the German forum of ProZ a few days ago wanting to know (among other things) if 70 euro cents per line is a good rate for EN>DE or FR>DE work. I do know people who have worked for less, but I responded rather bluntly that I would not have gotten out of bed for that 10 years ago, and I certainly would not today. She didn't say what field was involved (since the question was about agency rates that probably isn't terribly relevant), but in the area of banking and finance, for example, I haven't encountered a colleague anyone was willing to recommend who charges less than € 1.60 per line. Most charge well upwards of € 2.00 per line. Plus rush charges and other relevant surcharges, of course. Those too are actual rates.

So rather than obsess about how low one must go - or even how high one can go in a field in which one has no expertise - I'll choose to keep my blinders on and look straight ahead at what is possible and reasonable in my own areas. I talk to a lot of colleagues about a lot of issues, rates included, and even those who are closest in specialties or skills have unique characteristics that affect the rates that they can and should charge for the same service I give. We can give each other advice all week long, but what really matters is how each of us filters that advice and applies it, accepting responsibility and learning from the results. In the long run that will be the difference between success and failure, not what is charged per word today.