Podcast: Play in new window | Download Subscribe: Apple Podcasts | Android | Email | Google Podcasts | Stitcher | TuneIn | RSS In episode 75 of The Content Strategy Experts podcast, Elizabeth Patterson and Bill Swallow talk about how content reuse can help you save on your localization costs. Related links:  Reuse in DITA and beyond (podcast) Use cases for content reuse  Twitter handles: @PattersonScript @billswallow Transcript: Elizabeth Patterson:     Welcome to The Content Strategy Experts podcast brought to you by Scriptorium. Since 1997, Scriptorium has helped companies manage, structure, organize, and distribute content in an efficient way. In this episode, we talk about how content reuse can help you save on your localization costs. Hi, I’m Elizabeth Patterson. Bill Swallow:     And I’m Bill Swallow. EP:     And we’re going to dive in to talking about how content reuse can help you save on your localization costs. So I want to get started with just a really general question, and when we talk about reuse, what are we talking about? BS:     That’s a very good place to start. When we talk about reuse, what we’re not talking about is copying and pasting of content. You could think of that in terms of reuse, but it’s not really what we’re talking about here. When you copy and paste content, you’re essentially duplicating it and then need to manage it in multiple places. What we’re talking about is more intelligent reuse of content, so writing it once and using it by reference wherever you need to use it. So this way it’s only written once, and it’s used multiple times as needed. EP:     Great. And we have done a podcast and an additional blog post just solely on reuse, so I will link those in the show notes. But I want to dive into now looking more specifically at how content reuse, now that we’ve defined that, can help us save on localization costs. BS:     Well, generally speaking, reuse reduces the overall number of unique words that you are translating. By using intelligent reuse in your writing once, and using it multiple times by reference, you have the opportunity to choose pieces of content that you will author once and only once, and that content gets translated once and only once regardless of how many times it’s being used. If you copy paste, you can still see a savings if the wording that you’re using is one for one, so if it’s absolutely exact all the time. BS:     For example, I know Microsoft Word has an auto text feature, so you can throw a basic reusable component like a caution statement or some other boilerplate text, and you can use that to insert it every single time. That may save you a bit of time on the offering side and ensure that the text that you’re inserting is exact every single time. The only problem with that is that it is inserted as normal text every single time you insert it, so it does still increase the total amount of words that you need to send to the translator. It might be a 100% match, but they still have to do a check against it to make sure everything is fine. And the systems that they use will still count those words and say, “Yes, this is a 100% percent match.” But it’s still being counted as part of your incurred cost, because there’s something that’s going to the translator for them to see, even though there’s a match. BS:     And in some cases you may even get what they call an ICE match, or an in context exact match, on that text. So if you are using something like Microsoft’s Word’s feature, you can drop that text in every single time and you can get this, “Yes, it is a 100% match every single time it’s inserted.” And if it’s a full paragraph, it could be, “Yes, it’s a contextual perfect match. It’s a paragraph and it says the same exact thing.” But more times than not, when you talk about inserting strings of text that say the same thing over and over and over again, the context may shift depending on where you’re using that text. In which case then you get maybe a 100% match, which still requires some review, or you get what we call a fuzzy match, where if you happen to make an edit to that text that was inserted and copied and pasted it’s no longer 100%, and therefore the translator has more work to do. BS:     And there may be questions. This one has two words that are different from this other block of text. They say roughly about the same thing, should they be translated the same way or is there a reason why they’re different? That just slows down your translation process, it injects confusion, it injects questions that need to be mitigated and answered or you can then suddenly have a divergence in the translation where you shouldn’t have. The translator might’ve translated it two different ways because it used two different structures. BS:     So true reuse or intelligent reuse moves that out of the way by taking the text that is being reused every single time, putting it somewhere to the side, is translated separately, and then can be used as it needs to be used throughout whatever it is you’re writing. Your manual, your web content, whatever you need to. And there are plenty of tools that are out there that do this well. Two of which that come immediately to mind for desktop publishing based tools are FrameMaker and MadCap Flare. BS:     FrameMaker uses a series of conventions where they store content in chapter files, and those chapter files are assembled within a book file. And you can easily reuse an entire chapter in multiple different books just by linking to that chapter file from the book. You don’t have to rewrite the information, it’s not being copied and pasted. It’s a dynamic link that goes right to that file and pulls it into the book. BS:     FrameMaker also has text insets which function a little bit in the same way where you have a separate file that has a block of text and you can say, “Hey, go to this file, grab that text, and place it here.” And the smart thing about this is that when you do that in FrameMaker you are not creating an editable copy of that text. It is a reference to the file that contains the text, but you cannot modify it within the context of whatever it is you’re writing. It is uneditable. You can see it, you can read it, but it is uneditable, and you can’t modify it. BS:     The same goes for MadCap Flare, where you’re building things in a similar fashion. Where you’re grabbing individual files, and you’re putting them together in an order to create some kind of document or website or what have you. And MadCap Flare also has something similar to text insets, they call them snippets, and you are able to insert these snippets throughout your content. And those, again, are managed in a separate place, they’re written only once, and they are non-editable in the context of where you’re using them. They’re only there as a reference point. BS:     Now, these are great, however, you do have some concerns when you’re using these tools for localization purposes. They’re not inherently bad, but if you are looking to do a lot more with your content, let’s say you are styling your content very differently for different outputs or you’re creating the same type of output, but the styling is different. The text insets in such are in the snippets. They’re going to, I believe, carry a lot of the formatting information over with them however they’re formatted, where they’re stored. So it’s not incredibly ideal, but it does reduce the total number of words that you’re translating. BS:     When you move to something like XML, you have a bit more available to you because you have these conventions, but they’re built into a format of writing that does not have the formatting applied to the content. So it’s all text base and you can do quite a bit with organizing and reorganizing your content without having to worry about your headings being formatted one way or another. It’s all just plain text and the formatting is applied at the point you’re publishing. EP:     Right. So I think what we’re seeing here, obviously, is that there’s really one main way that you’re going to be saving money on your localization costs through reuse, and that’s just reducing that word count. But the way that you go about making that happen in your strategy is really going to vary depending on where you’re at as a company. BS:     Right. EP:     So I want to get into a few tips. So what are some tips that you have for reusing content, particularly when you are planning to localize that content? BS:     Well, the knee jerk response to anyone who is doing localization for the first time, and has all of this reuse potential in front of them, is to reuse as much as possible and to apply conditional text or conditional formatting as much as possible. And even I was guilty of that many, many, many years ago where we would have a manual that would go out in 19 or 20 different languages but one of them was over in Europe. And I figured, “Oh, well, for the English stuff we’ll just condition in and out the characters that differ between certain words. We’ll condition in or out a U in color or we’ll condition out a Z for an S for localize.” These types of things. And I thought I was being quite inventive and it came back immediately that no, you cannot do this, because when you send something for translation the translator gets a wall of garbage that they’re looking at and wondering what you’re trying to do with these words. BS:     So my first bit of advice is do not go too granular with your reuse. Things like reusing words or phrases, I would really limit that as much as possible. You really want to reuse at a larger chunk level. So if we’re talking DITA or if we’re talking something like MadCap Flare, reusing at a topic level. So here is a topic with a heading and a bunch of text or a procedure or what have you, reuse that whole piece. If you need to reuse it five, six, seven times, that’s great. You’ve written it once and you can leverage it an additional four or five times. That’s fantastic. Reusing things like notes, cautions, warnings, they tend to stand on their own. I mean, they’re used in context with other text, but the warning itself, you can write those to be very standalone as far as what the thing you should not do is and what the outcome of that is within the context of that warning statement. And you should be able to put that off to the side, write it once and use it everywhere. BS:     There are two benefits to that. One is the localization impact and the other one is that all your warning messages are exactly the same wording. And it will drill that information into your readers’ heads over time as they read it to say, “Oh yeah, I shouldn’t do this. I should not do this.” There are only so many ways that you really should say, “Don’t stick your hand in the machine while it’s working or you’ll lose it.” You really want to say it only once and repeat that statement multiple times until it’s drilled into your audience’s head to, “Hey, don’t stick your hand there.” EP:     Yep. And you want to make sure that it’s being said in the same way so they don’t take a different meaning from that. BS:     Exactly. Or have it translated differently even though you meant to say the same thing. EP:     Right. And something that I’m thinking about as we’re talking about what companies give their translators, so that their translators are trying to figure out what they mean, is writing style. So when you get into an organization that has many different writers, what are some things that you need to be aware of when you’re planning on sending content to translators? BS:     The first thing you have to do is have your style guide nailed down and make sure that all of your writers are following that guidance. Sometimes in larger organizations where you have too many authors, and perhaps not enough editors to clean up after them, you might want to look into some kind of editing based software or language based software like Acrolinx or Congree to do a lot of the spot checks automatically, rather than relying on someone to catch it in proofreading. Especially if you have tight timelines, quick turnarounds, and everyone’s just too busy to proof each other’s work. I know that the days of having a fleet of editors cleaning up after writers has kind of run its course. There are still many technical and editorial editors out there, but not to the degree they used to be in, let’s say, even the 1980s when, unfortunately, I started working. EP:     Right. And content governance can help with that as well, right? BS:     Oh, absolutely. The more you can nail things down and have a process for how you produce your content, the better off you’re going to be. And the one thing you absolutely must do, and I wanted to touch upon this also with the style guide, is you have to include your localization people in that overall plan for governance in styling as well. You want to bring them in to help define the language style that you’re going to be presenting this information in. The way they’re going to write their translations, how you want their translations to read, and which words they should use, and which words they should not use and why. BS:     You really need to have a global style guide at that point and be able to provide glossaries of information to your translators, because you may have different translators for each language depending on when they’re available to take on the work. Unless you’re fortunate enough to have them all in house as employees, which is extremely rare. A lot of times that work is outsourced, whether it’s through a language service provider or you’re doing it direct with other freelance translators. So being able to have that global style guide in place, to have a global glossary in place. And what’s really critical is being able to, when you do have these reusable components, that you’re going to be giving them to translate not only the components, but the content where the component is lacking. Because when you’re reusing by reference that content does not exist in the file that they’re looking at. BS:     So you want to be able to provide additional contextual information to the translator to say, “Oh, hey. When you get to this point there’s a bit of content that’s being inserted.” And maybe even provide them with the reasonable content to say, “This has already been translated, but this is what’s going in here.” So that way when they get to that point they’re not stumped and say, “Well, this doesn’t make sense because it goes from part A to part C, we’re missing part B. I don’t know what it says there.” That can certainly throw off the translation process. So being able to provide that additional context around what is going on in your content set is critical when you’re doing things with intelligent reuse. EP:     Right, right. So I think really one of our main takeaways from today is that you certainly can save money when it comes to localization on the translation side of things, but you need to be prepared to really pay attention to your translator’s needs. BS:     Absolutely. I mean, the savings that you get from a reduced word count is all fine and good, but the translation is only as good as the quality of the translation itself. And if you’re tripping up the translator in any way, you’re not going to see that return on an investment in localizing. EP:     Right. Absolutely. Well, I think that that’s a good place to wrap up. So thank you so much, Bill. BS:     Thank you. EP:     And thank you for listening to The Content Strategy Experts podcast brought to you by Scriptorium. For more information, visit scriptorium.com or check the show notes for relevant links.   Use cases for content reuse Reuse in DITA and beyond (podcast) The need for a localization strategy (podcast) A hierarchy of localization needs Localization strategy: not just cost minimization Localization strategy: Your key to global markets