Webinar

Break Your Sales Records by Using Localization

Matija, Črt and Adam discuss localization and what it is, why it is a crucial part of a business strategy when trying to reach new markets, and how it will break your sales records.
Native language ads will always have better click through rates because a larger portion of the population can understand it, some people are against English on principle, on nationalism, some people don’t speak it. So, automatically, it improves your click through rates when you are using a native language. ” – Črt Podlogar
Matija Kovac
Matija Kovač
Co-founder / Head of Development
@Taia Translations
Matija Kovač is the co-founder and head of development at Taia Translations, a company that bridges the gap between language barriers with the help of Artificial Intelligence and machine learning.
Kaja Sepec
Kaja Šepec
Human Resources Manager
@Taia Translations
Kaja Šepec is a psychologist and an HR expert who is responsible for pulling all the right strings at Taia, in order to create a tight-knit and warm culture, in which each idea counts, and where every voice is heard.
Crt Podlogar
Črt Podlogar
Ecommerce Marketing manager
@Sensilab
Črt Podlogar is an experienced digital strategist, consultant and speaker who has more than 10 years of international experience in website development, marketing, CRM/ERP integrations, analytics and e-commerce sales.
Adam Wooten
Adam Wooten
Co-founder @Acculing, Associate
professor @MIIS
Adam Wooten is the co-founder at Acculing and Associate Professor of Translation & Management at MIIS. With more than 10 years of experience in the industry across all of the translation areas,  Adam is considered one of the best in the industry.

What you'll learn

  • The differences between localization and translation, and how those two can affect your company’s image.
  • How you can improve your click through rate by using native language and break your sales records.
  • How to help local buyers find your website.
  • How to prevent international buyers from leaving your site before completing a purchase.
  • Everything you need to know when expanding your business into foreign markets.
  • Why Amazon, eBay and other big players are implementing neural machine translation as part of their localization strategy (and why you should do so as well).

Transcript

Kaja Šepec: 
Okay. Hi there everybody and welcome to the Taia webinar called Break your Sales Records by using Localization. I’m really happy that you’ve taken this time and decided to join us, and that you’re as eager as us to learn more about localization and how it can help e-commerce companies.

My name is Kaja. I’m the Operations Officer and HR manager at Taia, and I will be moderating this panel discussion with our amazing guest speakers. You will soon get a chance to meet them all, but first let’s plan the course ahead and let’s set some ground rules. So we’ll start off with a moderated panel discussion, and during this time you will be muted, as I mentioned before. So if you have any questions during this time, write them in the chat or in the Q&A section – it’s in the bar below – and we’ll make sure to answer them. However, at the end of the panel, there will be a whole dedicated part just to Q&A. So you’ll be able to ask your questions live.

In case that you’re not comfortable with that, there will still be an option to write your questions in the chat and in the Q&A. So, yeah, I think this is all, and without further ado, let’s go on and meet our speakers. I would ask Adam, maybe first, to start off and introduce yourself.

Adam Wooten: 
Hello, my name is Adam Wooten. I am a translation and localization consultant and executive, and I’ve previously started my own localization company, and I also teach in the Translation and Localization Management degree program at the Middlebury Institute of International Studies at Monterey, California.

Kaja Šepec: 
Thank you. Črt, could you follow up now and introduce yourself?

Črt Podlogar: 
Yeah, no problem. Hi, my name is Črt. I’m the E-commerce Marketing Manager at SENSILAB pharmaceuticals – let’s say a mid-sized food supplement and slimming supplement e-commerce company, mostly selling on European markets – about 80 markets, 80 web stores. And before that, I used to work in marketing as a consultant for a performance marketing agency for about five or six years. Before that, I worked on e-commerce projects in Studio Moderna, which is also an international e-commerce company – about slightly more markets than SENSILAB, about 22-23, mostly working with managing data, opening new web stores, launching new projects in different markets and things like that.

Kaja Šepec: 
Thank you. And last but not least, Matija.

Matija Kovač: 
Hi everyone. My name is Matija Kovač. I’m the co-founder and Development Manager here at TAIA translations. I’m actually a language enthusiast, studying Chinese and other languages during college times. But by heart I’m a developer, so now I’m leading a team of developers and we’re building this translation platform that helps companies localize and translate their content more effectively and easily than they could do it before. So welcome all. It’s nice to meet you.

Kaja Šepec: 
Alright, then. So let’s start off with questions. Adam, localization has become increasingly important in the last few years and it has taken on a role of kind of a buzzword, but there are still many who don’t know much about it. So for those who might not be that familiar with this term, and might be a bit confused by the concept, can you maybe talk a bit about what localization means and why is it important for a company which wants to expand abroad?

Adam Wooten: 
Certainly. So most people know what translation is, and I like to define localization in terms of translation as being not only translation, but localization is also adaptation, and that adaptation can be cultural, technical, or regulatory. And so there are many ways that you can adapt something in cultural, technical, or regulatory ways, and if you don’t, there can be a flop. So, for example, if you don’t adapt a product technically, to even be able to support the translation and localization, your translation might not even display properly, and then your e-commerce website or whatever product you’re putting out there can bomb in that way. If you don’t adapt it culturally, when necessary, you might not be able to make that connection with your users or your buyers that would be necessary to make a sale or have a successful product. And if you don’t adapt it in regulatory ways, say for example, complying with particular tax collection laws online, then you might even be banned from certain markets. So it is important to be aware of when that adaptation is necessary or advantageous so that you can localize your website or your product appropriately.

Kaja Šepec: 
Thank you. Črt, I think based on what Adam said, we can all agree that localization is a very wide field and it definitely isn’t just as simple as translating the context to the target language of the local market. It’s far more complex and more nuanced. So based on your experience, what are some aspects of localization that are worth being emphasised, since we’re at the point where we are defining it?

Črt Podlogar: 
So, I would even say maybe that localization — you know, when we’re talking about localization, we always think in terms of culture and language. I would even go one step further and maybe say that market specifics can impact localization as well. You know, in a sense of what we see in our business of ours, is that existing competitors or biggest players in our markets can sometimes really impact a couple of aspects of our e-commerce business, from USPs that we communicate with, from the type of ads that people are used to, the profiles that really work, or let’s say, in some cases, even design and structure of e-commerce websites that people are used to from the big players on the local markets. Once you start entering these kinds of markets, it’s really important to understand this as well and I think that, you know, I would put, I guess, all of these things into the term localization, and yeah, I think it’s really really important, and we often see that when you’re really looking at performance or when you’re really in this — you know, the classical e-commerce business is really EBITDA-oriented and usually lower average order values are lower prices, and high sales volumes usually mean that you really need to nail the profitability and conversion rates, and with these kinds of businesses it’s really really important. So, in my opinion, taking a look at competition and adapting to what the competition is doing is very important. Ads and communication, particularly on all of the platforms and channels; pricing, which can be reflected by both the market and the general buying power of an individual market; but also, the prices that the competitors set, the USPs that you’re going to use or develop, maybe USPs that you’re going to take from the market and start using yourself, or maybe USPs that you don’t see on the market, but you can deliver with your product or with your services, and you might want to try to use them and push them on the market because nobody else does, you know?

So the only thing that’s really a hard decision for us often is do we want to sing with the choir when it comes to certain aspects of localization, or do we want to differentiate ourselves and try to make that a key important thing. Payment systems, I think, are very important as well, because we all think about it. We need to provide a credit card, but when you start working on individual markets, sometimes to customers it’s important that the payment provider to whom they’re going to give the credit card information is actually a locally-recognised brand or an international brand that’s localized and has a presence there, so the type of payment system that you’re using can sometimes cause problems in terms of conversion rates. Certifications and trust badges – there’s a lot of companies that offer these kinds of certifications, trusted shop, safe shops, certified shop, or anything like that. We found that this has a very positive impact on sales as well, as long as you choose the locally relevant players, and one thing that Adam already mentioned, and I think is really important, is product compliance as well. We are a producer of most of the products that we sell and product compliance – not just product compliance, but especially, maybe in the food supplements business, it might be a bit specific, you know, I’m sure it’s different if you’re selling TVs or food supplements – but in our business, it’s also very important what kind of ingredients we use because, you know, you can achieve a specific effect with multiple ingredients, and it’s really important to pick the ingredients that are most recognised or most commonly used on specific markets. And this can vary quite a lot or in some markets, some ingredients might even not be compliant and you wouldn’t be able to sell the product. So it doesn’t just affect localization in a sense of what kind of content you prepare, which is what we usually think about, or what’s going to be on the website, but in some cases it can even impact product redevelopment. So it’s really important to start thinking about localization as early as possible.

Kaja Šepec: 
Thank you.

Adam Wooten: 
If that’s okay, I’d love to just comment as a quick follow-up. I love that what most of what you were commenting on, whether it’s with payment systems or product ingredients that are well-recognised, or even doing what people are used to seeing from the big players – so much of that comes down to making sure that you can be viewed as a trusted website or a trusted partner or supplier, and that is a big part of localization. You want to localize at least enough to where you can be trusted.

Kaja Šepec: 
Yeah, exactly. Thank you. So let’s move on to Matija. Matija, let’s imagine this company A, they have an amazing product, it’s selling great in their base home country and now they decided they want to expand abroad. They’ve already identified the target countries, but now they’re wondering whether to localize it or not. I mean, everybody speaks English anyway, right, so why do it? What does, in your opinion, a fully localized site tell to the buyer and what kind of effect might it have on his buyer behaviour?

Matija Kovač: 
So, yeah, this is, I think, something that both Adam and Črt already touched upon and I think Adam put it very wisely when he said that a certain aspect of localization is definitely the cultural side, so adapting not only the language, but also bringing it closer to the community that’s the target audience here. This is definitely very important, but I’ll get back to that in a second.

I think, first, what you need to be aware of is that about half of the internet currently is still in English, but the amount of content that’s being produced in other languages is growing exponentially, compared to English especially. And, only about 20% of all the internet users are native English speakers or, in a lot of cases, people don’t even speak English when they visit the worldwide web and we’re talking about users from not only European countries here, which are quite often not English speakers, but also users from many growing markets that are very interesting, especially for services that can sell and deliver also to these markets. If we think of the sizes of markets like China and Russia and India, most of these people have very poor levels of English or they don’t even speak it. And what’s interesting here is, even to visitors on your website who actually speak English, it’s the cultural side that gets them. I think it was Nelson Mandela who famously said – and this is something that’s been tossed around in the translation and localization industry for decades now – but he said, “if you speak to a man in the language that he understands, it goes to his brain, but if you speak to him in his mother tongue, whatever you’re saying goes to his heart”. So the emotional factor here is what’s driving this debate that I’m trying to put out here.

In fact, there was also this research done by the University of Chicago on the field of psycholinguistics, where they actually came out with the results that communicating in a foreign language, even if you understand it, takes out the emotion of your deliberation process, and in e-commerce this is extremely important because what e-commerce is very famous for is that people go about buying products sometimes in a very impulsive manner, and it’s hard to be impulsive if you have to translate everything again in your mind. So it’s much easier to get the user to buy impulsively if they have everything already in their native tongue, even if they speak English. And there was another research that came out with the results that about 75% of consumers rather buy in their native tongue, so, I think it’s quite important to understand this.

Of course, there’s another aspect that Črt also put out before about the price and cost effectiveness of this, so you probably won’t go about localizing your entire batch of 10,000 products or whatever you are selling in the first go. You will probably try to identify what are your key products or services that are most logical for the market that you’re trying to target here. And you’ll first go with localizing those, and then probably you will move forward, especially with professional human translation, which can increase the costs quite fast, in fact.

Kaja Šepec: 
Thank you.

Črt Podlogar: 
Maybe if I can just jump in…

Kaja Šepec: 
Yeah.

Matija Kovač: 
Please do.

Črt Podlogar: 
Just one quick comment, because, this whole psychology of impulsive buying in language and how it makes it easier. I perfectly agree, but there is also one thing that a lot of people forget which is purely mathematical, and that is, in most cases, at least throughout my career that I’ve seen, is when you compare English language ads, let’s say, specifically versus native language ads, native language ads will always have better click through rates because a larger portion of the population can understand it, some people are against English on principle, on nationalism, some people don’t speak it. So, automatically, it improves your click through rates when you are using a native language. And click through rates are one of the key ways how most buying algorithms determine what the price is for your clicks, conversions and everything else are going to be because if your click through rates are lower, that means that an ad platform is going to have to serve your ads more to get the clicks and more to get the conversions and purchase the purchases that you want, which means that you as an advertiser use up more of their advertising space that they could be using to get clicks and revenue from somebody else. Ergo, your cost of advertising will go up and the more, let’s say, native language-focused the market is, the bigger this extra cost you will be paying can be. So sometimes it can be, you know — it’s differences of sense that we’re talking about. But once you’re paying for thousands in clicks and hundreds of thousands of ad impressions, this can really add up to considerable numbers.

Matija Kovač: 
So the return on investment and properly localizing your ads, for starters, and then all the other content is definitely worth it in this case. It’s a solid point.

Kaja Šepec: 
Yeah, great. So, Adam, you’ve been in the localization industry for quite some time now, and I guess you’ve seen it change tremendously over the course of this time. Especially in the last years, there has been a big surge in the advancement of AI and neural machine translation. How do you see these new technologies impacting the localization process in the future and how can we fit them into a smart localization strategy today?

Adam Wooten: 
Yeah, certainly. So we do see that, particularly among the big players for the past decade. Amazon and eBay and others have been using neural machine translation and other forms of statistical machine translation as part of their localization strategy. So it’s something that’s here to stay. Now, of course it doesn’t always turn out exactly how they would like. We saw just last month when Amazon launched a Swedish website using a lot of machine translation, that they got a lot of bad press for it, because there were a lot of mistakes in it. But that doesn’t mean that they’re wrong to use machine translation at all. I’m sure some of you have seen rather – if you haven’t yet you can look that up in the news – and there are a lot of rather amusing or scary mistakes that came up with that.

But I think that, of course, one of the big mistakes that a lot of e-commerce sites will make is thinking, okay, does it have to be all or nothing? Do I have to machine translate everything or do I have to just go with human translation for everything? And sometimes when you think of machine translating, everything is okay, do I just use generic, Google Translate or something like that? And typically the right answer is not all or nothing, and we even see that from the big players, they aren’t doing all or nothing. They’re saying, hey, you know what? There are some instances where maybe this horrible machine translation is better than nothing because we have thousands or millions of product descriptions and we just — we aren’t yet able to get human translation for all of it because we don’t have that budget. So maybe a lousy, horrible machine translation is going to be better than nothing, but there’s some other things that we should really prioritise, particularly things like what Črt was just describing, like even search engine ads or whatever it may be, because you really want to get those click through rates to be very high because there’s a great cost associated with them, and you want to make sure that those are human translated and adapted very well so they have high click-through rates.

And then maybe you look at your products and do a lot of prioritisation and say, “Hey, you know what? These are the products that these other markets in other languages are tending to gravitate towards more. So let’s make sure that they have a better human translation”, while these others that, you know, people are only clicking on, you know, a few times every several months, maybe just a machine translation is good enough and of course this is going to differ even depending on the product. Some products require a much more emotional connection to cause them to make an impulse buy or whatever it may be, and so that might require more adaptation and more human involvement. Whereas other products might be very technical, and have no emotional connection whatsoever, and sometimes a horrible machine translation is enough. And then sometimes there are even products that are in between or that, maybe have a medium amount of attention from other markets or only require a medium amount of emotional connection or personalisation, and maybe you might even have some combination of machine translation that is then post-edited by human translators and post-editors.

And so it’s important to realise that there is a spectrum – a large spectrum – it’s not always all or nothing, but there are stages in between where you can have some human involvement, some machine translation involvement, and we’re going to see that machine translation is going to continue to improve. Every year, we’ve been seeing improvements. Even now, just this year, we’ve seen that Facebook launched a multilingual machine translation engine that can go between a hundred different languages without using English as a pivot language, which is a great advancement…

Matija Kovač: 
It’s an amazing advancement in the technology.

Adam Wooten: 
Yeah, it’s fantastic. So, we’re going to continue to see improvements, but even with all these improvements, we’re still going to see a spectrum of use. There are going to be times when it’s appropriate to use it, times when it’s not appropriate to use it, and those times in between, where we can use a little bit of it and still use humans. So it’s important to talk with those who understand machine translation about some of that strategy and I’ll touch on that more.

Matija Kovač: 
Yeah, I would love to touch on this as well. If I can extend a little bit, cause it’s –Adam opened the whole new frontier here about this discussion and I think it’s important to understand, as you already put it out, it’s not only whether yes or not to use machine translation. It’s also how to combine it with other services in the translation industry, with other solutions. So on the one hand, yes you could go with just Google translating everything, but that would probably result in quite poor quality of the content. But on the other hand, what you can do is you can select the few products that are most profitable for your business, and you can select the main pages, from a practical point of view, of your website and have those translated with machine translation with post-editing, which is mostly a standard these days, or at least it is what we do at Taia. And you would probably invest more so, even editing and proofreading for some of this content – that’s your cornerstone content as it’s sometimes known as in SEO, or your pillars – you would probably translate those with at least three humans involved.

But some other content, you can only translate with the very easy human overview only which is something that, at TAIA, we call machine translation with a light review. It’s basically a new product that wasn’t available, what, three years ago, when machine translation wasn’t of such high quality as it is today, because now you can have a human translator to just basically review and remove the most common mistakes or the most obvious mistakes from the text and still keep the price quite low compared to an entire TEP translation, editing and proofreading process.

And then the next layer that you can go about is, if you have enough data of your own and if you’ve invested some to human translations already, you can have a private machine translation engine exclusively for your e-commerce platform, for example. This is getting more and more common, but it’s still out of this world for some older players who are not used to having proprietary technology exclusively for them. In this case, what we would do is we would harvest all this data, retrain a neural network to provide custom results exclusively for your business, and in this case we’re not talking about Google Translate anymore. It’s not this goofy translation that might be full of errors all over the place. It’s actually something that’s hard to separate from an actual human translation in many cases, and it’s comes specifically in handy when we’re talking about things that are very repetitive, like product descriptions that are usually even generated from attributes of this product and so on and so forth. So it might be a very decent solution with a fraction of the cost in the long-term if you have a large volume of data to work with.

Adam Wooten: 
Yeah, custom training of a machine translation for a particular subject area, say for example, health supplements, can help to make sure that there’s more predictability so that, if for example, it is health supplements, you can make sure that there are particular ingredients that you get right every single time. You don’t have to worry about the machine guessing so much because you’ve trained it on that type of translation.

Matija Kovač: 
Yeah. Of course this comes into play with the players who have enough content already translated or have some previously translated existing content that they can put into good use here so as not to waste what they’ve invested into localization in the past.

Adam Wooten: 
Yeah, it’s something that requires a bit of investment that can really pay off.

Matija Kovač: 
Exactly, yeah.

Črt Podlogar: 
Maybe just one other thing to add is, because we’re in the process of exploring different options of localization right now as well, and I was working with other companies and other clients in the past that were in the same position, you know, and when it came to machine learning, it was always these comments about machine learning isn’t the most accurate thing. But then on the other hand, nobody really thinks that humans are the most accurate thing either. Especially if you’re a company that’s in a budgeting or cost optimisation position, then you have these discussions “no, no, no, forget about machine translations, let’s go on Fiverr and find some people who are going to translate these kind of things”. And in our business, when, exactly what you pointed out, you know, accurate translations of ingredients can sometimes be very important because of legal compliance, and if we mistranslate something, that can cost us a lot of thousands of euros if somebody finds that mistake. And in a world of cost optimisation, a lot of companies also choose to go in a very not optimized way, despite thinking we’ve done the right thing because we chose the human way to do it. But, you know, the human way isn’t the most accurate either. So just something to put out there.

Matija Kovač: 
Yeah, it can backfire quite easily and we’ve all seen it in the past. So yeah, what you pointed out makes perfect sense.

Kaja Šepec: 
Okay, thank you for this conversation. It was really nice, I think – an exchange of ideas. So let’s move on with our questions, and to Črt. So you have a very vast, on the field experience, Črt, with your current and past companies. So if we put a number on how many websites and online shops you manage and how many foreign markets, what’s the number?

Yeah, you’re muted.

Črt Podlogar: 
Yeah, I muted myself. Sorry for that. I think I stopped counting websites at around 40, and markets probably between 20 and 30, I guess. But I have been sticking mostly to our side of the Atlantic Ocean, so not too much of the Americas or Asia in terms of experience.

Kaja Šepec: 
Okay. So based on this experience which you’ve gathered with all of these websites, what are some things that companies need to do or take into consideration when they start to think about expanding abroad?

Črt Podlogar: 
So the one thing that I always try to figure out is potential size of the market, and I think this is the first thing — you know, the problem with expansion to other markets in my experience, what I’ve seen, and it’s this mistake I keep repeating throughout my career – I think I’ve seen a lot of other people make as well – is we always tend to judge everything by our cultural and, kind of, experiential scope, I guess, so it’s very easy to misjudge – you know, what we, or what a lot of companies typically do when they think about expansion to other markets is they take the local experience and they just apply it to the population size of Germany, let’s say. But you can’t just multiply the situation that you’re currently in on your local or home market, let’s say, and then just, in the case of Slovenia, judging based on the market developments, trends, tendencies, cultural experiences of 2 million and say “we’re moving to Germany and our market is going to expand from 2 million to 90 million”. I think that’s about the amount of people that Germany has. So when I tell people you need to judge the potential size of the market, I always tell them you need to judge the potential size of your ideal customers. You can go to a much bigger market than where you currently are but your actual market potential size there is going to be smaller because there is less interest for the products you’re selling, because there are more competitors, so it’s going to be a much smaller slice of the pie. You’re going to get compared to what you have at home. So the first thing is deciding where you want to expand in the first place and I think that trying to estimate this potential size is very very important.

How to do it is an entirely different thing and it’s not so easy, but what some of the things that we usually do is — I really love keyword search volume on Google as one sign of what’s happening in the market, and there’s usually two things that I check. I compare keyword search volume numbers between different markets to try to estimate. If we’re talking about, I don’t know, sales of some B2B equipment, the number of searches for the relevant questions or relevant technologies or competing software in market A and in market B, and the difference between them can give you an idea how these two markets compare in terms of the potential that these people are going to have. And another one is how the search trend has been evolving through time, let’s say. I usually like to check what has been happening on a market for the past five or 12 years, let’s say, just to see, you know, are we going towards a growth trend for whatever I’m trying to sell, are we actually already in the decrease or are we on a plateau where we can expect a lot of competition, a lot of costs and so on. So I think that these are some of the things that are the most important to decide. Market maturity is something that’s also kind of a buzz word, but I found it very important as well, because like market maturity — everything has a market maturity every niche. I usually tell people everything from toilet paper to food supplements to 65-inch K ultra HD televisions. Everything has a market maturity.

Everything has a start, it has a plateau and then it has an end where we’re going to switch off. It’s just that the toilet paper has a very long market lifetime cycle, I guess. But this market maturity usually defines a lot of the things. It defines how you need to communicate because in other markets when people aren’t really familiar with what you’re selling – this is especially true for startups – the right kind of communication is especially important. So localizing things and how you’re going to use language and how you’re going to present your products is much more important than when you’re talking about selling the television in the European Union, where we all know what 4k means, what UHD means, what NGR that means, so there’s not a lot of explanation. You can be very technical. You don’t have to have very long and very explanatory product descriptions. You don’t need a lot of supporting content and stuff like that. So market maturity can really impact how you go about going to other markets. I always check the types of content that already exist on the market and how much of this content already exists, how much media are writing about specific products or industries or categories, how many bloggers there are, how many people are posting stuff about this type of products on Instagram, on Facebook and so on – just how much market awareness and market generated content already exists and what is the level of this content, so that you can get kind of an idea of how you need to position your own content in this regard.

And then one thing that is kind of technical really, but I think a lot of people forget as well, is delivery costs and rejection projections. Especially with low average order values and big volumes, delivering to foreign markets is usually more expensive. If we’re talking about high volumes, it usually means, in some cases, local hubs because it’s going to be more efficient and stuff like that, and how many rejections you’re expecting, based on some industry standards, based on what information you can find from, let’s say, any kind of delivery partners that you reach out in local markets because delivery costs and rejections are usually what kill projects in the markets and companies, because they don’t expect to have such high rejection of aids, they don’t understand that in some markets it’s perfectly common to order three different sizes of a product and then returning two that don’t fit or ordering the same product from five different vendors and whoever delivers it first is the one you pay, and everybody else you reject, or you just send it back, or you don’t even pick up the order. So, you know, rejection rates can be insignificant – 1%, 2%, 3%, or they can be profit-crushing 15%, 20%, 25%, you know, and all of these things are usually market-specific so it’s really important to check them out and see whether or not you really want to compete on this market and whether or not you have the logistics that can handle these kinds of markets.

Kaja Šepec: 
Yeah, that’s perfect, thank you. So moving on, Matija, companies who decide on localization, they’re often complaining that the process is simply too time consuming and that it’s not feasible since there’s this daily inflow of new content that needs to be translated and, due to the limited resources, they just cannot spend that much time on that process. How do you propose they approach such a situation?

Matija Kovač: 
Well, I think, again it depends on the service providers you select because it was quite common – and it still is with about 80% of the market, I would guess – to have your texts that you need translated or localized sent via email to your service provider, in this case, a translation agency or a language service provider, or however you want to call them. And they would usually need some time to process this order or this quote and send it back to you. It takes time for the project managers to manually handle these files and send you the estimations and yeah, we ping pong with emails back and forth before you would finally settle on the service that you actually need for this because you probably weren’t specific enough from the start. And it takes some time, especially if you don’t have the same LSP, language service provider, all the time. What’s happened in the last couple of years is there’s been a surge of new translation management solutions similar to what we’re doing here at Taia, where you can just basically drag and drop your content into the app and have it analysed in seconds. So what used to take a couple of hours or even up to a day or two for some project managers to send you back an email with a quote, is something that we can handle in about 30 seconds and entirely automatically in the middle of the night if it’s necessary. And then once you have the basic quote, you can just select the type of service that actually fits your needs and your budgets, and this makes it much more convenient for the users to visualise what the costs associated with translating some content will be, and also to understand how this will impact the quality, because in the good old days, as some people might call them, you would always pay for the translation with editing and proofreading, which is not something you might need in all cases, especially when we’re discussing e-commerce. Just as we said before, translating a description of a laptop does not require a proofreader. You can have it done mostly automatically and a human can just verify that the Machine translation was correct.

And if you go even a step beyond that, there’s all these integrations with e-commerce platforms via API protocols, so applications talking to one another, that if done properly – and again there’s a minimum upfront investment to handle this – but you can have a system ready where your content managers, like Črt probably has them in their company, would just select something inside their existing CMS or e-commerce, the platform that they’re managing, and this would automatically be sent over for a translation to their selected service provider. And it can either be only a machine translation or a custom machine translation, as I’ve already mentioned previously, or it can be any level of human involvement required that they might see fit. So with these API integrations, we came into an era that we now call continuous localization, so something that’s being localized as it’s being created. And there are some good use cases, for example, for this, like a good case story we had just a couple of days back in the company when we were discussing about this is Booking.com, for example, the traveling giant company. They were one of the first to implement their private machine translation because they have so much user-generated content that they need translated instantly and it’s better to have it translated instantly on a machine translation that is trained with their certain data than to not have it translated at all. And this is a similar approach where you can also involve an external LSP instead of just a machine translation and have human translators translate something without you even doing anything at all, once this is set up and running. So basically you would just click a button or flag it in your CMS and the content would be translated and imported back into your CMS whenever it’s ready, with the quality that you desire. So this is the beauty of connected systems that we can have today with the technology.

Kaja Šepec: 
Perfect. Thank you. So we are entering the last 15 minutes. So I would like now, all the participants, to invite you to ask questions if you would like to know anything from our speakers. If you want to ask live, I have opened this section, and just raise a hand. I will give you a chance to unmute yourself and ask your questions or do it in the chat or in the Q&A section.

Prashanth, I’m trying to pronounce this correctly, ask away. I’m asking…

Matija Kovač: 
You’re muted, you should unmute your…

Kaja Šepec: 
I’m asking you to unmute now.

Prashanth: 
Can you hear like me?

Kaja Šepec: 
Yes.

Matija Kovač: 
Yeah, we can hear you well.

Prashanth: 
Yeah, yeah, yeah. It’s Prashanth. I’m very sorry, I missed a lot of the seminar. I stay in India, you know, it’s a different time zone and I only get like a few minutes to spend with my parents for dinner time.

Matija Kovač: 
No problem, we will have a recording available for your Prashanth.

Prashanth: 
Yeah, I will use it. I even used that OTT Seminar. But then I wanted to ask you, is there any good platforms for subtitling that I can deal with my clients and I can use and again, promote it to my clients and I can work on projects?

Matija Kovač: 
So to understand the question correctly, Prashanth, you would like to translate the subtitles yourself, or would you like someone else to do it for you so you have subtitles already available online.

Prashanth: 
So, I’m a project manager, so my career wish is to be a good project manager and to grow in that field. So most of my clients ask me do you suggest a platform or do your translators have a platform? So, as you know, translators use many platforms. Like, Trados is a bit complicated, but there are many others on the field. So what would you suggest? So that’s what I want to know.

Matija Kovač: 
Unfortunately, I cannot help you. I see it that one of the users has posted a link here in the comments section but I think…

Adam Wooten: 
Diego.

Matija Kovač: 
[unintelligible 43:49]…might have some solutions.

Adam Wooten: 
Yeah, Diego. Diego’s posted some and really there are a lot of subtitling tools out there, even a lot of free ones like Aegisub, A-E-G-I-S-U-B, and many others and, so yeah, there are a lot of options out there. I would recommend that you experiment with them. I wouldn’t say that there is one clear subtitling tool that is just head and shoulders above the rest and the absolute best, but from what I’ve seen, they all have their strengths and their weaknesses and it’s good to experiment.

Prashanth: 
Yeah, I don’t know. Like, a lot of platforms have their own plus or minus points, but then the firm which I got into the industry had a lot of plus points but they couldn’t succeed in the field – maybe because of a lot of personal issues within the firm or the leadership issues, but then, it was a very good platform. I hope I find something like that. So thank you. Thank you. Thanks a lot.

Matija Kovač: 
Thank you for the question, Prashanth.

Kaja Šepec: 
Perfect. Any other questions from anyone else? Just feel free to raise the hand or ask the question in the Q&A or in the chat. Also Diego responded and said he agrees that there are many, and this one is just the one that came to his mind as he last heard of it and because it’s free. Okay, we’ll wait a bit still, but then if not, we still have some questions and 10 more minutes, so let’s see how much we can squeeze in. Okay, I’ll proceed with asking the questions and if anybody still wants to ask anything, just feel free. I’ll keep my eye out so we won’t miss it. Good.

Adam, based on all that we have said until now, it’s quite obvious that when we want to enter a new market as an e-commerce company, it’s really important that we build a solid localization strategy. So based on your experience, how should the company go about forming one and what are some steps they need to take along the way?

Adam Wooten: 
Yeah, well I think — I hope that based on what the panelists have said, particularly what, Črt said toward the end there, it’s important that we recognise that there’s a lot that can go into a localization strategy and it would be absurd to think that there’s any one person that has all the answers. I think it’s important to recognise that no single person can have all the answers and so, as you’re putting together your strategy, you need to reach out to others. You need to reach out to localization experts, but localization experts also need to recognise, hey yeah, we aren’t going to be the only ones that have all the answers. You need to reach out to your other departments and marketing, your teammates who are experts in data analytics, and you need to reach out to those in specific markets who understand users, and the users themselves. So as you’re going out, forming a strategy, recognise that you couldn’t possibly have all the answers, especially if you’re going into a lot of markets.

It’s hard enough to really become an expert in your own market, but to become an expert in 20 or more is, especially with so much that is changing, as Črt was talking about, even cultural or technical issues that can cause buyer behaviour and user behaviour to change even over the course of a couple of months. It’s important to recognise that, yeah, you can’t possibly have all the answers and recognise that once you have a strategy in place, things are going to need to change with time. However, as part of that strategy, recognise also that there are ways to prepare and make localization faster, cheaper and easier. Often in this industry, we refer to that as internationalisation, and that can hit all of the points of localization that I mentioned before – the cultural, technical, and regulatory or compliance-related issues. There are ways that you can internationalise your website and internationalise your approach so that localization is faster, cheaper and easier. And some websites are going to allow, and some e-commerce sites are going to allow, for more internationalisation, more standardisation and neutralisation, particularly those that are more technical, that don’t require as much of a personal connection.

But doing that type of preparation, again, can make things faster and cheaper and easier, whereas others will really just require heavy localization and not as much preparation will help you out. But yeah, as you prepare, for example, on the technical end, you may want to talk to your localization department or your localization vendor about internationalisation testing, using a simulation like pseudo-localization to be able to say, hey, are there any internationalisation bugs that could cause some major technical issues that might slow us down or really dramatically increase our costs or our turnaround times. If you do a simulation with this fake pseudo-translation, just to create a simulated translation of a copy of your website, that can really help, very easily and very visually, to identify what potential problems there might be, so you can avoid those problems altogether before you even start and not be surprised by them at the end and wonder why you’re going to miss some of your deadlines right before the holiday season.

Matija Kovač: 
If I can chip in here, this is something that’s quite common for website localization because you would have a website that’s in one language and when translating it to another language, the other language, let’s say German for example, might be even up to thirty or more percent longer in content and it might break your website entirely. So if you were translating, let’s say from Chinese, where you have about 10 characters for an entire sentence, into German where you have 250 characters for one sentence, in some cases, you would do well to at least try with just the machine translation just to see what the output might look like before you venture into actually investing large sums of money into human localization and ending up with a website that’s broken in multiple parts.

Adam Wooten: 
Yeah, that’s an excellent example. Another good example to consider is if you’re going in from a left to right language, as we’re speaking now, English, into a right to left language, like Arabic, Hebrew…

Matija Kovač: 
The design breaks completely yeah.

Adam Wooten: 
Yeah. The design can break completely or you might think that it works and in reality it doesn’t just because we’re not accustomed to looking at things from right to left, whereas it would be incredibly obvious to somebody who’s grown up as a native speaker of Farsi or Arabic or Hebrew or something like that. And so that’s another reason why it’s important to check with users who are accustomed to those markets and languages.

Kaja Šepec: 
Okay. We still have some questions. I’ll ask one and then we’ll have to find a way to deal with other ones. I think it’s the best that then we give access to all of the speakers, to all the participants, and they get the chance to ask the questions offline because we just have five minutes left. So, Marina was asking us if there are any courses or postgraduate programmes on localization or translation that you guys would recommend. She is originally from Croatia, but would consider online remote options since here there are not that many opportunities for this. Yeah, well of course I recommend my own programme, the translation and localization management Master’s degree at the Middlebury Institute of International Studies at Monterey, California. It is primarily on site, at least when we’re not going through COVID, but we do have an online version of the programme, that is this year, in the first year of its launch, it’s available to students in the US but we hope to open that up soon to international students as well. And so, yeah that’s one that I would definitely recommend.

Matija Kovač: 
I can second that. I’ve had the privilege to meet some of Adam’s students and I was incredibly impressed by the level of knowledge and engagement they’ve shown. So, yeah, I think it must have been a pretty good course since the students were really amazing. So please continue, Adam. I’m sorry for interrupting you.

Adam Wooten: 
Thanks for your support in echoing that, Diego as well in the comments and, yeah, so that’s great. Unfortunately, some of the other online programs that I’ve been aware of, like there was one at the University of Limerick that unfortunately is no longer around. That was available online. And so we do see some that are trying to pop up, but I hope we’ll see even more. You know, we don’t mind there being other programmes as well. I think that I would love to see localization programmes, even, you know, undergraduate and postgraduate, in just about every country. But right now, yeah. Oh yeah, Diego points out there is a short certificate, of course, from the University of Washington that you can get. There’s a free version, but if you want the actual certificate, you can pay for that. It’s a very short, three course certificate and in fact, many of our students have completed that certificate before they come to us and they come even a little bit better prepared and sometimes students who do that and do a little bit of study on their own, are even able to shorten their coursework when they then enrol in the Middlebury Institute. We do have a number of more experienced localization managers who will complete their degree with us in only one year instead of two, for example.

Kaja Šepec: 
Okay, great. Thank you all. So maybe since we are not able to answer all the questions, where can the people reach all of you guys to ask some further questions? We’ll share those details with the email with the replay with everybody, so they can ask ask their questions.

Matija Kovač: 
I’m always available on LinkedIn, for example, so that’s a good channel to get me over.

Adam Wooten: 
Likewise.

Črt Podlogar: 
Hear, hear.

Kaja Šepec: 
Okay, good. So we’re going to make sure that people get those details, and yeah, I would like to thank all of you, Adam and Črt and also Matija, for participating in this great conversation and also thanks to all the participants for listening in on us rambling about localization. So yeah. Thank you for your time and have a pleasant evening, all.

Matija Kovač: 
Thank you everyone. Bye.

Liked this content?

Get notified when we publish something similar.

* We don’t spam your email or share it with anyone!