Need for translation management, small budgets, low-tech
After some years of working as a freelance German to English translator for a university, I did a survey among colleagues about the structure of our niche of the language services industry. These colleagues may be on the buyer-side, like full-time translation managers, academics with a part-time role in translation, or freelancers.
First of all, thanks to all respondents. Your insights are valuable for language professionals in our niche and beyond. In summary, there are some special features to the university translation market:
There are many ways translators can use their experience to create value for institutions, beyond just translating texts. For example, most jobs come as Word of PDF files and only basic computer-assisted translation (CAT) features are needed, so university translators can help save money by investigating the available tools
One advantage of having your users come to you is that you can get quick feedback to decide which content will actually be read. In my experience, organizations should know exactly what they’re publishing and what gets translated, otherwise they risk leaking money into automated processes to translate content that nobody uses. Collecting this info is another way in which steady translators can assist content creators.
To start with a comment I received, I hope that despite fairly low budgets, this survey can help not just to raise the prioritization of translation, but also align the priorities of translation professionals with those of the institutions they serve:
“Our main problem is that we have neither the necessary staff nor the necessary authority to truly assume all translation tasks for the University, resulting in a wild mix of styles and degrees of international accessibility. It is simply not a high priority for the University headship to speak about a consistent international strategy that actually includes a language and translation strategy.”
Over a third of respondents work at universities with more than 10,000 students. These are likely to be traditional universities (Universitäten), the smaller ones are more likely to be so-called applied science universities (Fachhochschulen, FHs). Germany has a problem of shrinking universities, so attracting international students can be vital.
Overall, (only) around 11.1% of all students in Germany come from another country. Most respondents to this survey worked at institutions within that range, but some had more. How the translation budget contributes to attracting international students and improving their experience at the German host institutions should be studied further. It is known that nearly half of all international bachelor’s students don’t finish – lack of support and social integration have been cited as reasons. Translation should definitely not be the only support internationals receive, but language access is part of the experience.
It’s no surprise that international students are the main audience for translated material. This question was intended to find out if the university attracts more international researchers or students. From my experience, professionals working in Germany need to know the language well enough to get by without translation, so it can be assumed that most texts we translate target students, but more data is needed.
As in my case, most internal customers at universities are administrators, marketing staff and program coordinators. These are core functions, so to speak. Researchers seeking to translate publications or lecturers trying to translate course material usually have to find external funding or do it themselves. Please comment on your experiences!
Most respondents have less than 1 full-time employees handling translations for the entire university. 28% have at least one full-time translator, while 16% have more than two. Part-time translation staff might be researchers with another focus, who take this administrative position to supplement their income, or they can be freelance translators who do this as a steady side gig.
In-house university translators often develop into a project managers, taking care of glossaries and translation tools and liaising with agencies and freelancers, but the translation workload often pushes management tasks to the side:
“At our (small) university the realisation that translations are necessary (for a university which, at least in parts, prides itself as being international) only occurred a few years ago, we have two members of staff each holding about a 25% translating position. Any formalisation of processes happens (if it happens) if we can fit the initiative for such steps into our translation schedules – so other that the day to day business of translations there is very little knowledge of skills and tools…”
Surprisingly, English is not the only target language, some institutions also translate into German. One even translates into French. Personally, I have never translated into German for my university, and I don’t know any use cases, but I assume most English to German material would be translated in-house.
University translation is a low-budget affair – most spend less than €10,000 p.a., which means a single freelancer would not be able to live off one university, but working for multiple universities could be a worthwhile specialization.
Considering that public sector contracting often requires good knowledge of the institution and communication with internal buyers, university clients could be difficult to bundle, and I don’t (yet) know any agencies or freelancers who do, but please leave a comment if you know better.
Interestingly, most translations at universities seem to be done in-house, either by dedicated translators or university staff in other roles. Since professional fluency in the target language is widespread in universities, there is less need for outsourcing of translations or proofreading.
In my experience, deadlines also work differently in the public sector – a translation is either so urgent that it is translated by the author, or it is has so much time that the project does not need to be split and the translators can set their own deadlines. The procurement process also tends to stretch deadlines, since translators may not be able to start working until they have a confirmation from central purchasing, and there might be several feedback cycles after delivery.
Since few translations seem to be outsourced, procurement is also very ad hoc. Not all universities have framework agreements, meaning they outsource regularly to the same providers and all orders are signed off by the same office.
One issue with centralized purchasing is that these roles may order anything from pencils to construction services. Very long and high-level RFPs are often the result, where it’s not clear whether the university is trying to build and manage its own freelancer pool or find an agency.
An issue with decentralized translations is that academic writers, for example, might just hire someone they know instead of going through the university’s “official” translator. This might make sense if field-specific expertise is required, while administrative translations should be more centralized, since this makes the language more consistent and shortens communication chains.
One respondent seemed to confirm the need for more specialized purchasing for language services:
“A lot of the questions about finances can only be answered by the budget manager, which is not necessarily the translator. Also, because outsourced translations are managed centrally, I have no overview of how these operate. University translators are not necessarily budget holders…”
CAT tools not fully leveraged
Interestingly, 64% of the universities use a CAT tool, but almost 90% do not use match discounts. The reason could be that most translations are handled in-house, so discounts don’t matter. From what I’ve heard, most universities use Trados; if you know better, please leave a comment!
I believe there is some savings potential here, since universities don’t need the full feature package. Most files are in Word or PDF format (see chart below), and often there’s no separation between project managers and translators. Furthermore, universities only need translation memories (TM), term bases (TB) and machine translation (MT), but not extended functions, like integration with content management systems or team collaboration features.
While Trados has a home advantage in Germany, translators should still invest in learning about features and the market for CAT tools, since technical consulting is one of the major ways in which we can contribute value by preventing vendor lock-in and combining best-of-breed solutions.
One respondent commented:
“Purchasing translation software is difficult, so the process is still fairly manual. There is a need for continuous localization, i.e. a CAT tool that integrates with university’s CMS, but this might require specialized consulting. [The] university doesn’t have a big picture overview of all translation/editing work, although there is a framework agreement with some freelancers.”
There are many cloud CAT tools that offer the needed functions at affordable subscription rates. I personally use SmartCat, and it does the job for free. My university also has a need for continuous localization, and SmartCat would integrate with my university’s CMS, but setting up the process seems beyond the scope of what the CMS manager is currently able to provide.
Continuous localization is coming, but there are still some obstacles. Translators who keep this conversation alive may find opportunities in the future.
Almost half of the respondents split urgent translations occasionally and 28% don’t know. The issue with splitting is that it requires strict terminology management and translation memory maintenance, and often there’s no one responsible for this. If these linguistic assets are poorly maintained and translators don’t regularly align with project managers and authors, inconsistent output is the usual result. Just like continuous localization, the conversation about crowd collaboration needs to keep moving.
Only 56% know for sure that their glossary is used by all translators. (This does not mean that the glossary is well maintained or that there is a process for vetting new terms.) In my experience, even if universities invest the effort to create a glossary, it quickly gets outdated. If there is no CAT tool or other software, the glossary may be difficult to access. If a glossary isn’t integrated into the translator’s tool, it won’t get used much. So having a glossary and having translators use it remain challenging…
Proofreading leads into the issue of pricing. While the ISO 17100:2015 standard outlines a workflow, in reality translations often are not proofread by a separate person, since a second pair of eyes costs money. Only 8% of respondents make this investment.
With target language expertise being common at universities, it’s no surprise that university staff do over half of the proofreading. 32% use the “single source” approach and rely on the translator to do the proofreading, which can work well, especially if the translator is familiar with the institution and knows whom to contact in case of queries.
The German Normzeile (55 characters with spaces) is used widely by public sector buyers. Yet, 25% of respondents pay by the word, presumably because they work with an agency. Other respondents don’t outsource, and hence don’t have a unit-based costing approach. I’d like to read your comments about outsourcing and internal costing of translations.
Surprisingly, 8% of respondents paid less than €1.00 per Normzeile. 25% paid between €1.30 and €1.60 and none paid as much as the €1.95 proposed in the latest revision of JVEG, the law governing compensation of translators and interpreters in the justice sector. Further differentiation would be needed here. As always, I need to emphasize that the bottom line depends on more than the contentious unit price. I would assume that universities with relatively large budgets are able to negotiate lower unit prices, but please tell me more!
Only 3 respondents charged by the word, with the lower range being between 10 and 15 euro cents and the higher range between 15 and 25. Around a quarter of respondents paid different prices for different types of content, a strategy on which I received the following comment:
“We pay different prices to different translators. Depending where they are located, VAT varies, too. Our head of team administers the overall budget for translations, and I have no idea how high (or low) it might be.”
Language service professionals working in and for universities should focus on technological opportunities, in addition to their language and process standardization efforts. There are many ways in which in-house translators can help maximize the amount of work not done, instead of just processing more and more text. They don’t just help select vendors for language services, but also for software.
As a freelancer working closely with a university, I am able to help in some of these areas, but lacking integration with the organization sometimes makes it difficult to recommend improvements. In my contract, I’m paid to sell words, not consulting services, so any advice I give is basically pro bono, and I’m not able to help with implementation or data maintenance.
In the future, I see translators working more closely with content management, marketing and procurement to help universities make money, save money, attract students and keep a better overview of the information they process daily.