This time we’re discussing the criteria a business can use to determine if having the capacity and resources to do translations in-house makes good sense.
Ian:Perhaps we can begin with a definition of what in-house translation entails.
Tim:In-house translation refers to either having your own staff carry out translations for you or potentially using an agent, distributor or reseller in a target market, which is something you see done regularly in the business of medical devices, for example. In both of these scenarios in-house translation involves having your translation done at ‘no cost’, using resources that you have access to get translations done rather than sending them out to a professional translation company.
In-house translation is often employed at the start of an organisation’s localisation journey. People think along the lines of “I know such and such who can speak French or German, so I’ll just get her to translate for us”, and in principle there’s absolutely nothing wrong with doing this.
There are various pitfalls however that are part and parcel of this approach. The most obvious stumbling block people encounter is that they underestimate the amount of resource that’s required to get a translation done. To put some rough metrics around this point, as I’ve outlined before the productivity that you can expect from a freelance translator is in the neighbourhood of 2,000 – 3,000 words per day.
Ian:I imagine this was a very common inclination in the pre-digital era when there was far less demand for online copy.
Tim:That’s exactly right. But while a professional translator can translate 2 – 3,000 words a day (which isn’t a particularly large translation job) you can’t expect your staff member to do this in a day on top of her day job. Realistically then a job that might take a professional freelancer a day to do might take your staff member 1 – 2 weeks to complete.
You might use a member of staff or a reseller. In both scenarios you encounter the same problem: neither you nor your resellers have people sitting around waiting to do your translations, so timeline is a major practical consideration.
Whatever size of translation you undertake to do in-house, there’s likely to be a significant impact on the project timeline when you’re utilising an ‘existing’ resource whether that’s a staff member or a reseller in your target market. In both scenarios you encounter the same problem: neither you nor your resellers have people sitting around waiting to do your translations, so timeline is a major practical consideration.
Also, as soon as you’re into translating five plus languages, co-ordinating that level of production is difficult. Your French and German translations might arrive on time but you’re waiting forever for your Japanese version, so when you have a product launch that requires all the translations being completed, you’ll end up going at the speed of the slowest translation.
In this scenario you might end up going for a mix and match approach, doing the French and German translations in-house and contracting the Asian languages out. It’s important not to dismiss the option of doing translations in-house because if you do have competent resources with available time, why wouldn’t you utilise that?
Ian:Is the task of in-house translation included in certain job specs more often nowadays because of the increased need for multilingual content?
Tim:What you’re saying makes sense but in practice I think it’s very unusual. On the whole people still think they can get someone to do this as a ‘favour’, and the person who’s asked agrees to do the translation with the caveat that she/he isn’t sure when it’ll get done.
Another thing to think about is the seniority level of the person you’re asking to translate for you, because sometimes very senior people carry out in-house translations. You should ask yourself in that situation why you’re paying this senior person in the region of £60-80,000 a year as she should have a high-powered intensive job. If she is translating for you, what is she not getting the time to do? Again, I’m not against in-house translation, but it only makes sense to use more junior staff who have the available capacity.
There’s a hidden overhead with doing multilingual translation which is something that’s often overlooked first time around. People assume that translating is in fact straightforward and not exceedingly difficult and often they have good relationships with people that they can ask to do it. The reality is that these processes take a considerable amount of management.
If you’re in a situation where there is a constant flow of copy that needs to be translated into one language you may want to consider hiring a full-time in-house translator, which is a much more rational decision. If your #1 market is Germany and you’re constantly producing technical documents for Germany, then get an in-house German translator.
Ian:We’ve been talking so far about in-house scenarios for smaller businesses with relatively modest translation needs, but how or when should larger businesses with substantial translation needs consider the efficacy of in-house translation?
Tim:What a larger company needs to think about before considering undertaking in-house translation is the technology and processes that underpin successful localisation, which is a combination of two things. Firstly workflow, so you have some kind of proper idea about who’s doing what and when you expect people to deliver, and what processes you expect a translation to go through such as whether a second person is proofreading as well as whether someone is doing document preparation. These are the things we do on a day to day basis and trying to do them piecemeal using Excel spreadsheets just doesn’t work on any kind of scale.
You need to have the resources available and you have to be certain that you’re being sensible in allocating the resources that you’re using to translate. You also need to have the technology to underpin your efforts and ensure you’re using your in-house resource efficiently.
The second thing is that any translation company will use CAT (Computer Assisted Translation) software tools, and it’s really important if you’re going to start doing in-house translation on any scale that you have access to those tools.
CAT is essentially a database of your existing translations. Every single time we translate something we transform it into something called a translatable file which goes through a process of parsing the text. You parse the text by extracting it from its original format and splitting it into what we call segments, which is typically a sentence but can be part of a sentence depending on how you segment any text. Instead of a Word document with paragraphs you create a series of parsed segments of text down the left hand side of a document with your translations of the segments facing these on the right hand side. You can then see each segment and its translation into the target language.
This goes into a database and the next time you translate that segment the software will alert you that there is an existing translation in the database so you don’t have to translate it again. This is both a major time saver in getting a translation done and a major cost saver when you’re getting an external translation done. What typically happens is that companies who are doing in-house translation (and most tend to do some) do so with no discipline and without the use of translation memory – unlike the external agency which utilises highly developed workflow processes and CAT software.
Ian:So to summarise, having the right software is a major necessity to order to undertake in-house translation efficiently on a significant scale?
Tim:The key message is that if you’re going to do in-house translation you need the right tools so that your in-house people are getting the benefit of and contributing to that translation memory and your termbase if you have one as well. You have to have the resources available and you have to be certain that you’re being sensible in allocating the resources that you’re using to translate. You have to have the technology to underpin your efforts and ensure you’re using your in-house resource efficiently.
Ian:What are some of the key criteria for a company to use when considering whether to make a commitment to the human and tech resources in order to undertake translation in-house on a significant scale?
Tim:If you’re going build in-house translation from the ground up of course you have to think about volume – do I have enough volume to in effect create a mini translation agency? Some bigger players do exactly that.
If you’re looking at a spend of €/£/$500,000 and upwards you’re at the stage where going in-house might be a financially rational decision, and if you’re at that level of translation spending you probably need to have an underpinning of technology. Even if at that level you’re still putting out all your translation through an agency you need a way of controlling this as you might be splitting your business between multiple agencies.
We offer our customers the ability to work on our platform, a way of translating in-house and getting access to the tools and processes you need without making the up-front capital investment.
It’s more challenging when you have a lower level of spending; if you’re spending say $100,000 a year it’s difficult to justify the investment in the software and cloud servers as well as dedicated staff to run the process for you. We offer our customers the ability to work on our platform, a way of translating in-house and getting access to the tools and processes you need without making the up-front capital investment. This allows the customer to interweave their in-house translations and what he sends to us or other translation services providers.
This option allows your in-house translators access to the correct tools and the right work flow process, and when dealing with difficult web file formats and DTP formats we can handle this so you haven’t got staff struggling to make things work that they’re not experienced with. In this way you can still do in-house translation if you’re not big enough to justify buying software and recruiting a full time team.