Mar 10, 2011

Cheap translation, like coffee

Coffee is a commodity. Except when it isn't. A recent set of articles in the New York Times explored the economics and problems of the international coffee trade. Like many agricultural commodities, coffee is subject to boom/bust cycles in which high demand leads to overproduction, which causes prices to drop to unsustainable levels, with production capacities eventually diminishing again through neglect. The overproduction, unsustainability and crash part may sound familiar to some in the translation industry, though the campesinos struggling in the fields of words seem to have little hope for boom years again.

But Übersetzung ist nicht gleich Übersetzung, and the same is true for coffee. There is a culture of specialty coffee, which emphasizes balance and quality and is fit for the purpose of drinking. One author making a case for drinking better coffee wrote:
"Here’s my advice: stop drinking cheap coffee. It’s getting worse, as those who deal in bargain-basement coffees blend in more and more of the 'bad stuff' to meet a price point. The “good stuff” will get better and better, as food lovers recognize the miracle that great coffee is, and value it as such. It’s a truism in food: eating less but better leads to maximum health, happiness and value."
This is rather good advice for translation consumers as well. If cost is really an issue, commission less translation work, but consider carefully the content to be translated and look for a real professional to do it to protect your image and the health of your business. The hash pipe dream peddled by the MT vendors, who tell you they can translate millions of articles with "acceptable" quality, requiring just a bit of post-edit tweaking by the campesinos who work for beans, is an amusing, but unrealistic hallucination that can take you on a very bad trip. But what of the masses of bilinguals ready to take on big volume projects assigned in the care of major, lion-hearted LSPs promising a single point of contact, top quality, fast service and low cost? Equally bogus. In some submarkets of translation there are surely some good translators willing to scratch in the dirt like chickens for a few grains to survive on, but any purchasing agent willing to risk is employer's health on this unsustainable situation had better have a Plan B for when it doesn't work out and he gets fired for incompetence. While price does not necessarily correlate with value, other issues such as timing, short lines of communication, availability and trust are usually - among intelligent beings - more important than price. Or would you rather send a highly confidential employment contract to an industry "leader" who sends out copies by unsecured e-mail to 400 translators around the world asking for their best price? You would? You're fired!

Like the producers of cheap coffee, major LSPs doing hundreds of thousands of words in a matter of days pour anything into the mix to meet a deadline and a price point. The only thing that usually gets left out is the quality needed to produce an usable result. To those who find this situation adequate, I would like to point out greater potential savings with much less hassle: Google Translate is free. Of course it's crap, but if you think what you'll get from a language service provider paying translators one to three cents a word is any better, it's time to stop hitting the pipe.

Of course it's difficult to generalize about the diverse world of text to be translated. A recent discussion on another blog about whether freelance translators earn "enough" discussed some of the differences between the axis of FIGES and those translating in more exotic language pairs. I live in the "GE" part of FIGES, so my perspectives on the world are inevitably influenced by this. However, certain basic principles do seem to apply across languages and borders even when the absolute numbers differ.

"Ah," you say, "But sometimes that cheap work really is good, even better!" True. One of my best "competitors", a better translator in my best specialties with more experience as a translator and a better work ethic, charged significantly less than I did in most cases for many years. She was the preferred choice for many projects and rightly so: I would even recommend that clients call her first and contact me if she had no capacity (often that was why they were calling in the first place). Well, eventually this person woke up to the reality of the market, and from our last chat a few weeks ago I gather that she often charges significantly more than I do and still has her capacity booked out. She just needed to learn to take the "risk" of asking for more and justifying it.

So I do not believe that it is a sustainable strategy to rely on cheap and good to stay that way. And those who are addicted to the low price of translation brew will inevitably end up sipping bitter dregs if they don't consider carefully what they really need from the experience of language transformation and focus on the real costs of getting it wrong.


  1. In some parts of the world there are many people who can't tell the difference between good coffee and pretty bad coffee. I am possibly one of them. At my local Food Lion or Farm Fresh supermarket, I have a choice between a house brand, which costs about two dollars less, and the Starbucks brand or something like that.

    I buy the house brand most of the time because both brands taste very similar.

    As long as people can tell the difference between real translation and garbage produced by MT, edited MT and subprime translators, real translators should be in good shape.

    I for one think that unlike with different coffee brands, it is quite easy to tell the difference between real translation and the fake stuff.

    With translation, I think that most people can tell the difference quite easily. And smart customers will hopefully realize that the cheap stuff is too expensive (car Le bon marché coûte cher, comme on dit on Français).

  2. I quite agree.

    In fact, if you underprice good quality work, your clients may assume low confidence in your abilities. Which is probably true. Clients often assume low confidence means low actual ability, but in an industry like ours - where we make our own success rather than being given promotions - that may be difficult for some.

    When setting rates, a translator should consider things like their client acceptance rate, their client retention rate, their client comments, their client engagement, what peers say about their work (in linguistic and business terms), their unique industry expertise in their specialisms, as well as their various signs of quality (memberships and, if applicable, ProZ KudoZ score and - somewhat more importantly - answer acceptance rate.

  3. I posted this on TranslatorsCafé.com :)

  4. "...people who can't tell the difference between good coffee and pretty bad coffee. [...] I have a choice between a house brand, which costs about two dollars less, and the Starbucks brand."

    Your comparison is faulty: Starbucks is to good coffee as Transperfect to good translation.


Notice to spammers: your locations are being traced and fed to the recreational target list for my new line of chemical weapon drones :-)