L10N Reality Check: Industry Insider Shines Light On The Dirty Little Secrets Of The Translation And Localization Industry That You Won’t Learn At A Webinar

By the Masked Localization Insider, special to The Content Wrangler

If you’ve attended webinars on globalization, translation or localization lately you’ve undoubtedly heard a lot of happy talk about the “revolution” occurring in the translation world. Cloud computing, crowd-sourcing, new aggregates of translation memory, mystic project management portals and constrained (or simplified) English are all supposedly leading us to a shiny new world where “invisible servants” translate and disseminate our content to a vast and disperse global audience. The 1939 New York World’s Fair promised “The World of Tomorrow” too. But there was also a huge pavilion for the asbestos industry. Television was prominently featured, it only became widely available 15 years later. Although World’s Fair bit the dust by the early 1970s, the same gee whizmentality persists when it comes to translation and globalization webinars. On a fundamental level, nearly all translation agencies are somewhat alike. So we can forgive the bit of star dust and show biz production that vendors use to help them differentiate themselves from the competition — to stand out from the crowd.

Technically speaking, almost everything you hear while attending a localization or translation webinar is true. But that “tomorrow” they talk about may be a little further away than you would suppose. In fact, you may find that your budget, resources, available internal workflow, and manpower may make you a bit like a customer in 1932 with his or her nose pressed against the glass of a Duesenberg show room. Yes, somebodyis going to buy that car. But it won’t be you. And by the time you are upgrading to a dramatically better solution, you may find that the “car names” and solution features have changed a lot.
Don’t get me wrong: “going global” to reach rich off-shore markets is a necessity, and translation/globalization is worth every penny that you pay. But it isn’t cheap and it often isn’t pretty. And yes, there are a handful of vendors out there who will take advantage of your ignorance and pass on unnecessary costs. Read on to find out how to stretch your globalization dollars, and be as effective as possible with the staff and content creators you have at hand. This last factor, your staff, (and out-of-date authoring software) is a huge “gating” issue that is seldom acknowledged by translation vendors during webinars or before, during or after a sale.

A dream world, still out of reach for most of us

Webinars by localization vendors promise a lot, especially a better tomorrow. Vision of clean XML and DITA files with total separation of content and formatting are described, with content assets housed in “chunks” on sophisticated content management system(CMS) servers that can quickly identify only changed portions of revised content. Improved software solutions would extract only the changed content and (via automatic email notification) send those portions, with reference materials to the translation vendor. Webinars also create the impression that your translation vendor has a Star Trek worthy project management “portal” that automates virtually everything.

All of these solutions do indeed exist, but there are also hidden price tags and limitations within your own workflow and headcount that will probably keep that magic at arm’s reach. When venturing into your first translation project, you will probably encounter unexpected costs with the software/solution price tag, additional post-kick-off services and also with your own internal resource costs or manpower hours.

Localization vendors are good people. But like anyone else in sales, they want to close the deal. So there is always the temptation to lo-ball actual costs when providing you, as a new potential customer, with your first project quote. A few translation vendors who will deliberately shave costs and “under bid” to get your business will usually nickel-and-dime you to death with change orders once your project begins.

The challenge of finding the right vendor

The translation industry is in the midst of considerable turmoil right now due to a series of acquisitions, dividing vendors between the large “box store” mega-stars, stable mid-tier agencies, and thousands of small (and sometimes highly effective) vendors. You really can’t judge how well a vendor will suit your needs based on:

• sophistication of a vendor website
• the number of overseas offices
• that project management portal personalized “just for you”
• or any number of other common factors

The best way to choose a vendor is still through word-of-mouth, social networking and direct contact with credible customer references.

Fortunately, in today’s world of Twitter, FaceBook, Linkedin and a slew of other socially-enabled networks, it is easier to do a background check than at any point in history. When considering a vendor, for instance, you can check the company on Linked In and count the number of former employees to determine the rate of turn over.

Here are some things to watch for.

Red flag #1: If you visit a company website and find that the CEO has his/her fingerprints and personality on practically every page (e.g. videos where the CEO takes up 90% of talking head time, all blogs written by CEO, or coy press releases that quote him/her a lot), run, don’t walk, to find any other vendor. This is a sure sign of a “closed” environment where creativity is percolating in slow motion.

Red flag #2: If the CEO and headquarters are based in Europe, and the company’s website doesn’t make it clear that your US division has fairly independent management, this can be another inhibiting factor. I speak from experience, having seen many a brilliant software solution shot down by a British-accented voice on conference speaker phone that asked “Sorry chap, but has anyone else done this before?”

Years ago, some people got on wagon trains to find a better world. And when the gold ran out, they invented the Silicon Valley. Other people’s ancestors were a bit less adventuresome. Just a thought.

Some other concerns to be aware of: The time zone differences between you and Mr. Big, and the plethora of UK Bank holidays may also crimp your project’s progress if the domestic crew isn’t fairly autonomous. Don’t get me wrong, not all UK-based companies are bad news. But I’ve personally had more than one experience with UK-vendors who simply did not understand the time metrics and dynamics of American business. Take away message: With any translation vendor, clearly identify all members of your team and what time zone they are in. Remember, if it’s bedtime for you Thursday night, you’ve just lost all of Friday in Asia, and you’ll wake up as they’re heading out for the weekend.

Bigger isn’t always better

One might be tempted to choose a larger translation vendor, thinking that dozens of overseas office, decades of experience and a large employee head count would guarantee project success. Ironically, larger, more conservative agencies sometimes stifle or discourage inter-department creativity. It really depends on the division you are working with and if you have a stable team of project managers, etc.

Some huge translation vendors have large enclaves of in-house translators, hundreds of desktop publishing specialists, but those resources are available on a “first come, first serve” basis. And then there are the 30 to 40 multi-country holidays that can make some of those off-shore resources unavailable at inconvenient times.

On a positive note, if a large translation company has acquired a lot of small to mid-sized companies, and had the brains to keep the original innovative managers on board, it could be a terrific partner for you. If you find that most of the US-based managers have foreign accents, this probably isn’t the case.

In some cases, you may find that a mid-sized, US -based agency with highly talented staff that has long tenure are more nimble in crafting a solution that will suit your needs. Some “big box” vendors won’t talk to you unless your starting out with at least a $50K to $75K project. In some cases, even a very small, “boutique” agency that has a highly focused niche (e.g. in vitro medical devices) can be the best choice.

Will XML and DITA “change your life” and raise the dead?

Nearly every localization and content management webinar will praise the benefits of the Darwin Information Typing Architecture (DITA) and and Extensible Markup Language (XML), combined with intelligent content management which promises to dramatically lower translation costs with eye-popping ROI. Under the right circumstances, all of this is true, but here are the limitations.

• There are a lot of formatting and document constructs that DITA does not support very well. Be prepared to simplify your document formatting and let go of some features (including sentence structure) that you’re in the habit of using.
• Be prepared for a fairly hefty price tag for an underlying database if your projects are huge and you have to exchange assets throughout a large company. Good news: there are some new DITA/XML-friendly content management systems that work well for small work groups in the US$15K to US$35K range.
• Converting content to DITA/XML isn’t easy or fun, and you will need at least one full-time, dedicated expert/maven on staff who can fully understand all of that underlying “hidden” structure. While there are many authoring solutions that will “hide” DITA from you during content creation, don’t give in to the temptation of letting a vendor own your DITA Sysadmin duties.
• Finally, make sure that your internal content creators and tech writers have “bought in” to the switch to DITA. If they aren’t happy with the process, their resistance can cost a lot of money and delay that return on investment for months or even years.

Take away message: Yes, DITA works, it is worth it, but it will cost more and be way less fun than you may be led to believe.

What’s wrong with Content Management?

Content management systems (CMS) are great solutions, and will probably eventually become mandatory for nearly all of us. Here is the biggest mistake that customers make: After attending dozens of webinars and getting “sold” on the benefits of DITA/CMS, some poor souls select a CMS solution without fully researching their own needs. This is sad because our industry content management gurus have been warning us to avoid this type of tools-before-strategy approach for years. They even tell us what we need to do to be successful — needs analysis, modeling, training in structured authoring/writing for reuse, change control planning, etc. — and we don’t listen. So we make the same expensive mistakes as the people who went down this path before us.

I worked with one client who was sold a bloated, highly customized CMS solution that came with hundreds of “surprise” consulting hours. Costs over two years exceeded $400,000 when production finally went live; today they could have reached their goals for about $50,000 with an “out-of-the-box” solution.

Worst mistake: some ill-advised customers let their translation agency provide the consulting for selecting a CMS solution. Yes, translation and content management are closely related, but not that related! Letting your translation vendor select your CMS solution is a bit like going to an interior decorator when you really need a plumber.

New Trend: Due to a series of acquisitions by larger translation companies, more than one vendor now offers its own XML CMS solution. Nearly all of these solutions are good and can produce great results, but beware; the software and resources required may be bigger and more capable than you need. Often, the acquired CMS solution was developed by another company at a time that translation and localization was not a core service. Just remember that whatever the job title of your localization pre-sales contact (Globalization Strategist?) you are basically dealing with a sales person. If they can add $60K to the sales ticket, they will.

Who “owns” Translation Memory (TM)?

For years, this has been one of the most misunderstood concepts by customers with translation projects. Translation memory is nothing more than an organized database of all of your previously translated text strings (software) or sentences and paragraphs (web and documents.) Linguistic software will “match” new source text to previously translated text and prompt the linguist to select an “exact” or “fuzzy” match to avoid “retranslating” the same text.

You have paid for every word of your TM and it belongs to you. It should be delivered to you, at least quarterly, if not upon project end, so you can use it with other vendors if you wish. There are a handful of unethical translation vendors out there who will “hang on” to your TM, and refuse to deliver it to you. They obviously think that you are chained to them for life if you can’t take your assets elsewhere. If this happens to you, calculate the word count of your TM, multiply it by your vendors per-word translation rate, and deduct the total from your next project invoice.

Why is DTP necessary?

Many customers new to translation are shocked that a substantial part of the quote may be taken up by desktop publishing (DTP). Why is that necessary? Basically, post-translation DTP is “document reconstruction.” Language text expansion (often word and line count can increase by up to 35%) can wreak havoc with page layout, but so can the linguistic software. Older legacy documents (Word, FrameMaker, InDesign) that are riddled with word-level format overrides can almost “decompose” when going through the translation process. Much manual DTP time can be required. And as one who’s been there, post-linguistic DTP clean up on “ugly” documents is about as refreshing as using cuneiform clay tablets instead of an iPad.

In the “perfect world” of a DITA/XML/CMS publishing and translation model, DTP time can be virtually eliminated. But the reality is that most customers simply don’t have the bandwidth to “scrub” their content for unnecessary formatting before translation. And this leaves a “cash cow” opportunity for a handful of unsavory translation varmints.

DTP specialists run wild!

DTP can actually be a litmus test for vendor integrity. Competent, ethical vendors will always do pre-project document analysis and deliver a laundry list of undesirable document elements that should be purged before translation, and even give you the option of cleaning up legacy source files yourself to save money. A handful of unethical vendors will greedily rub their hands together, leave you in the dark, and happily bill you for hundreds of avoidable DTP hours. When DTP costs approach 35% of overall project costs, be suspicious.

What about those vendors with really “cheap” DTP rates?

Some translation vendors have attempted to be more competitive by moving all DTP staff off-shore, primarily to China, the Philippines and India. This enables them to charge customers as little as $10 an hour for this service (vs. $35 to $45). Have you ever bought one of those shirts made in Bangladesh and had half the buttons fall off after the first laundering? Well, you get what you pay for, and post-translation DTP isn’t very different. The “sweat shop” methods used for some offshore DTP stables do not bode well for quality results. And internecine communications between two managers who barely speak English can lead to further errors.

At the risk of sounding politically incorrect, I’ve personally observed that DTP workers in mainland China have not been encouraged to suggest innovative improvements — or to question obviously “incorrect” instructions from authority figures. If instructions don’t explicitly spell out what can and cannot be done, problems may ensue. For instance, safety regulations in China prohibit use of toxic chemicals in paint (e.g. Barbie’s lipstick), but not in “jewels” glued onto toy bracelets. (Remember the recent Walmart incident?) The cultural trend in China is to follow directions to the letter, even when the directions are obviously wrong. This is a very general observation, but in my experience it happens more times than not when dealing with post-translation DTP instruction misunderstandings.

Take away message: if your product has anything to do with saving lives, get references for DTP quality if Asian off-shore resources are used. South American and east European DTP pools will cost a little more, but the quality is substantially better, and you will sleep a lot better at night.

Customer desktop publishing staff party like it’s 2002!

And before we leave the subject of DTP, let me comment on one of the biggest mistakes that customers make: not upgrading their authoring and content creation software. This may come as a shock to the folks at Adobe and Microsoft, but in a recent informal telephone survey I conducted with 40 other localization salt mine workers, I discovered the following:

• 75% of our FrameMaker users are still on Version 7.2 (release in 2002, has no UNICODE support, requires major hand DTP clean up in eastern European languages)
• About half of our customers are on Adobe Creative Suite 3 (CS3), they never upgraded to CS4, let alone the recently released CS5
• And, surprise, surprise, anyone who can is using Microsoft Word 2003 because they absolutely detest the user interface in Word 2007. Sadly, Word 2007 has underpinnings of XML that make DITA authoring and CMS practically “magic” with the right third party software.

Why are so many customers living in the past?

• Management won’t provide budget for the right software tools
• Fear of change: They are just “used to working this way” (exact words from dozens of clients)
• Most commonly, many clients have offshore divisions with obsolete hardware or operating systems and are unable to upgrade their software. So the US division publishes to the lowest common denominator. This basically blackballs you from that “World’s Fair” solution you’ve been hearing about. If this crowd wants to see the Trylon and the Perisphere pictured at the beginning of this article, they’ll need a helicopter, because the software they are using isn’t going let them through the gates of “The World of Tomorrow”.

Takeaway message: Software vendors, hire more software solution evangelists. It is a full-time job for each major product line. And no, you can’t have one person effectively promote FrameMaker, RoboHelp, TechComm Suite, eLearning and God knows what else. Don’t let the myth of “increased employee productivity” with post layoff/attrition headcount leave you with a molding customer base.

Why do I have to pay for Project Management time?

If I could name the one, most contentious issue in our industry (when it comes to invoicing), this is it. Many customers feel like they’re being screwed when their projects have a 10% to 13% project management fee added to the total. It feels a bit like hotel room service with an 18% gratuity. What if you don’t like to tip?

Project Management is a necessary evil with translation projects, at least for the foreseeable future. No matter how “magic” an automated project management portal is, there is still a huge amount of traffic to manage for any sizable project. And that includes a lot of email, and “interpreting” customer instructions into very simple English that off shore linguists (who barely speak English) can understand.

Which vendor has the best linguists?

In the 1990s, it was still fairly common for even mid-sized translation agencies to have a number of in-house, dedicated linguists. With an ever tightening economy and customers’ unwillingness to pay for “down time,” nearly all linguists worldwide were “let go” and they all became independent contractors.

What does this mean? There is a huge pool of off-shore linguists out there, and competing translation vendors frequently use the same vendors for different clients. Is this bad? Not really. Linguists are carefully screened, highly qualified (usually on some professional level, e.g. banking vs. medical device), and are bound by highly legal non-disclosure agreements.

What this does mean is that all translation vendors have to be highly creative to differentiate themselves from the competition. Most vendors have white papers, blogs and other ephemera on their corporate websites to establish credibility and special talents. One of the best ways to judge a vendors credibility is to attend several webinars and ask some really tough questions.

Closing advice

Obviously, I haven’t painted a pretty picture. And since I’m the masked localization/salt mine worker, you can’t send me an email to ask for advice. So here are some closing suggestions if you are either new to translation or considering the big leap of faith to a higher solution:

• To test new vendor competence, have the vendor perform at least a modest pilot project test. You may have to pay something for this (which is only fair). The usual practice is that the modest cost of the pilot project is deducted from the full project if you award the contract to that vendor.
• If your vendor has a substantial portion of your project team off shore, get out a world clock and count up the time zones. Also be sensitive to the number of secular and religious holidays (especially in the Muslim world) that will create pockets of down time during your project.
• Use social networking to find contacts who have worked with or even for certain translation vendors. Use Linked In to search for companies and note the turn over in staff. Feel free to contact a couple of former employees for an unvarnished view of the vendor’s practices.
• When you attend conferences or webinars, realized that the presenting vendor is probably a well-meaning, professional SME, who strongly believes in his/her solution. But all vendors (and linguists) live somewhat in a bubble; we work with the latest technology (when possible) and, unlike you, we never create original content.
• Follow some industry SMEs like Ann Rockley, Don DePalma, Tex Texin and Yves Savourel. Also join social networking subgroups that specialize in localization, translation, globalization and internationalization, where you can post questions and help set priorities.
• Most important, remember that you are the SME for content creation, not the vendor. The vendor (especially CMS or for authoring software) only “plays” with the software; they don’t really put original thought to pixels for life-saving purposes like you do.And so, let the buyer beware. And may the best vendor and solution win your business.

  About the author

  

  The Masked Localization Insider

  This article was written by the “masked localization insider”. He describes himself as a “salt mine worker” who has been burrowed in the translation industry for over 10 years. He doesn’t speak at conferences or write blogs; as he puts it, “I just do my job.” He attests to having worn a lot of hats, from sales, production, even engineering, for several domestic and European translation agencies. Since he likes his job, he is keeping his identity secret to avoid reprisals from outraged localization vendors, (especially the British and the Chinese). He claims to have more to tell, so if you like what you read, let The Content Wrangler know, and there may be a sequel.

Similar Posts:

None Found