Transcription
Max Morkovkin 00:06
So it's a full time and after the ball, we'll be having our next speaker who's Jorge Russo, the Santas based in Redmond, Washington. Jorge has a very interesting experience working as localization manager at companies like Facebook, Microsoft, and Cisco. And currently Currently he is responsible for localization of technical recommendation and training for Amazon physical stores, and also to teaching localization and the the University of Washington. And the presentation will be about selecting a localization model and a localization vendor. Jorge with us, okay. Hello, I see you.
Jorge Russo dos Santos 00:48
Hi. I'm good. It's actually George because I'm originally from Portugal. So we pronounced Ajay like in English, not like in Spanish. I'm sorry. I didn't hear no, no, no, no worries. I appreciate the effort. I always like when people call me Jorge. I feel special. So looking forward to it. So whenever we come back, I'm here.
Max Morkovkin 01:09
Okay. Okay, George. If we would have time, I would read again, everything. But I think yeah, you're given a thank you. So I think it's time for for the poll. Right? You asked us to do some poll at the beginning. My colleagues will share the link to mentee and the attendees, please follow the link, use the password. And we're looking forward to see some insights, right client or provider. And here are the options localization buyer, localization service provider, localization service, individual or company. Okay, I hope it helps you.
Jorge Russo dos Santos 01:53
If it does, it's just to kind of get a little bit more information of what people want to get from the presentation. So thanks, everyone for filling it in. It's nice to see the results popping up.
Max Morkovkin 02:11
Good, interesting. Are you well prepared for sharing the screen once the poll is over? Yes. Okay, good. How long time we'll give it 1520 seconds. So sounds good. Ah, there is another one. What is your role? And how many? Three
Jorge Russo dos Santos 02:33
said there are three questions in total?
Max Morkovkin 02:36
Okay, okay, good. How's it going so far, when you joined wasn't just now or you had a chance to listen to some other presenters,
Jorge Russo dos Santos 02:47
because I'm on the West Coast of the US. So it's morning now. So I just joined for the last presentation. But I hope to go back and review the rest of that content later on today.
Max Morkovkin 03:02
Yeah, and we will publish recordings, so you can watch them later for sure.
Jorge Russo dos Santos 03:09
Exactly. What's important to that?
Max Morkovkin 03:11
Would they want to learn? Oh, out of curiosity? What is the like, I thought that you were teaching localization in University of Washington? What is the length of these localization course? Is it half of the year? Or is it like longer or shorter?
Jorge Russo dos Santos 03:28
So that I ditched the localization Project Management module of localization certificate at the University of Washington. So the molecule takes three quarters, so nine months out of the year, but the localization game part takes one quarter or three months. So it's normally either between April and June, are between July and September, and there's like, only online even. So talking before COVID There is a module that is only online and is one that is either in presence or online.
Max Morkovkin 04:10
You've been working in many big companies, do you help best students to join the conversation to other companies,
Jorge Russo dos Santos 04:19
I give them tips to you know, go through the interview process because it's it's obvious that if I'm working in this company, and I have to go through that through the interview process, and I refer some of them when they when they asked me to do it for for some roles. So fortunately, I have a couple of success cases that I can refer to.
Max Morkovkin 04:49
Great great okay, then it's your time and I will turn on the video. Enjoy.
Jorge Russo dos Santos 04:55
Thank you, and you guys can see my screen. I'm I'm assuming you Yes, Yes, true. So thanks, everyone. Welcome. Wherever you are in the world, it's it's morning for me. But I realize it's evening and one of the last sessions for many of you. So thanks for sticking with it. And then thanks for doing the poll, because I kind of wanted to do this as much as possible, as a conversation where you, you get what you need from it, because it's not going to be an exact science where, you know, you find out exactly what is the localization modeling the localization vendor that you want at the end of this, but it's more like, what are the best ways to find out what works for you and your, your company. So let's get started. So, I'm going to do a small introduction, just talk a little bit about myself so that if you have any questions later on, you can refer to to my background. Then we're going to talk about how to select the localization model. And then we're going to talk about how to select a localization vendor. So as I mentioned before, my name is George, I'm originally from Portugal. And that's why my name is pronounced George in Korea. If I was from Brazil, people would call me Georgie, because they liked the vowels, much more than we liked them in Portugal, we kind of tried to compress them while they tried to open them up in, in Brazil. So I been working in localization for quite some time, actually, since last century, kind of started, like a lot of people started in localization. By accident, I started out as localizer for Microsoft localizing the first version of the Windows operating system into European, Portuguese, in Dublin, Ireland, then I kind of evolved in my career, I moved to the West Coast of the US to the Seattle in 2003. And I kept working at Microsoft until 2015. And since then, I've been working at other technology companies, like Cisco, workday, Facebook, and now I'm working at Amazon. I'm also an instructor at the University of Washington for the localization certificate, specifically in project management in localization. And I also authored one of the edX, which is like, an online course on localization. So check those out, if you are interested. And here is my information on email and LinkedIn if you guys have any questions or want to connect. One thing, though, is that these opinions are my own, you know, they don't reflect my current or past employers. So take this with the, with a grain of salt, it's, it's based on on my experience, and just take them as my own. So having said that caveat, so we had a couple of questions, what you want to learn the reasons for attending if you're on the client side, or on the provider side, and the role, so thanks for filling that out, it was good to kind of get an idea of, of the audience. So now we're going to talk about the localization model. So typically, this is separated into in house model where you have everyone under the same roof, maybe now, with COVID. You know, it's the same company, but people are under different roofs. But it means that you have a team that is all under the same organization, that there's another different model where you have an minimal or a skeleton crew, with the client, and then you rely on the community to provide your localization and translation needs. So we'll talk about that and how you, you can leverage that for your organization, if appropriate. And then we're going to talk about multi language vendors. So this is a vendor or provider that gives you an offering of many languages. So some of the advantages of this is that you can push some of the engineering and project management tasks into your render. So you can have a smaller team or you can dedicate place team to working in localization on the client side. So we'll we'll talk about this and how you can use this for your needs. And then we're going to talk about single language vendors. So in case it's not obvious, the main difference is that between the MLB and SLV, is that in this case, you're gonna have one vendor per language. So there isn't much that you can do in terms of shifting project management or engineering into the single language vendor, because you might have different levels of skill and expertise across vendors, but you are still outsourcing your translation and localization needs to an external provider. So these are the most common models, I'm sure that some of you are thinking that, Oh, I use a different model. And I would appreciate if you put this in the questions or the chat, and then we can talk about it better on and actually, there's kind of a bonus model that we're gonna talk about, that I haven't referenced here, because I think it's a bit of them have two of these. So let's start by talking about the in house model. So when, when you have an in house models, some of the greater advantages of this is that you have a lot of control over what you do. So you are, you have the whole team together, you can share communication, if there is something that you need to pass on in terms of information changes, you know, over time, things like that, it's very easy for you to control all of those things. So you also have more flexibility. And you also can achieve a lot more quality, because you can control who you are hiring, but also because you can control the flow of communication. So you can have your developers and your translators working side by side, if there are any queries, you know, people can talk about it directly, they can see the product, the in, in a metaphorical sense, they can touch it, they can feel it, they have a lot more information when you are all under the the same organization. So there are a lot of advantages to having a localization model that is in house and most companies that want to do localization, try to start off this way, because it's much easier to see the progress you you have a lot more control over all aspects of the team. So it's, it's a really good way to start. And also, if you're working on something that is confidential, and where you have to control all of the process, you don't want any leaks coming out either in terms of the product or the features, it's it's a really good model, because you have a lot more control over what comes out. So we talked about what are the advantages of this model? So let's talk a little bit about what are the disadvantages of working on this model. So the cost, it's, it's gonna cost you a lot to have a team in place, you know, you have to have the the office space, you have to go out and hire people. Depending on where your team is located. You might have to pay for relocation costs, you might have to deal with, you know, work authorization, issues, and all of that. So this is not for the faint of heart unless you are like working in big metropolis, let's say London, where you have people from all over the world, it's it's going to cost you to to do the same even in London, you know, you have to pay the market rate. So it's not going to be something that is for the faint of heart. Hiring is going to be a problem. You know, you have to temper your expectations because a lot of times when people are hiring, you know, they put out a job description, where do they want to have everything so they basically want to have top shelf software developer, that is also top shelf linguists that are So understands about, you know, machine translation and your ethics and, and all of that you kind of have to find the balance on the skills that you need and how much you are willing to pay for your team. And then you pretty soon you're gonna run into an issue that comes up is that you start off small, the localizing just a little bit of, of content, and then you start to go through the first hiccups, which normally happen, you know, when someone moves on in their role, or they need to take an extended holiday for whatever reason, or when you hit a certain threshold of volume, and then all of a sudden, you start thinking, Well, I have my team here, but they're not keeping up with, with the volume of work that is necessary. So how can I do this too, I can just expand the size of my team, how am I going to deal with this and also, like, as you know, there are discrepancies on on the productivity across languages. So you know, you might have a translator that has a really big output for, let's say, Dutch, and then you have translators that has a smaller output for the Japanese. And so you have to understand that those things are normal, it's based on on the language. So you might need for instance, to localize for Japanese while you only have one for that. So it's something that sooner or later is going to come up, which is like, what are the the issues with the volume? And how much can we do it from an in house perspective. And so it comes to the issue of scalable, so maybe you can do all of the localization in house for all of your software needs. But then when it comes to content that is related to user assistance, user education, marketing, it starts to get too much. And you have to decide if you want to keep growing your team, or if you need to look into a different model. So where is this a good option, I think this is a great option, when you are setting up a process. And you want to make sure that it's high quality. So in in the beginning stages, if you are now in a startup, or if you are in in a new division of your company, and you know that you're going to have a really strong international presence, this might be a good way to set up everything. And maybe you don't do all of the languages in house, maybe you just do a couple of pilot languages in in house. But it's it's a good way to have a lot of control over the quality, the communication, the confidentiality. Also, if you're talking about really specialized content. So for instance, like let's just say that SpaceX wants to open up their travel in space, for other markets, it might be really specialized content, it might be confidential, so it might be a good idea to start off with in house model. And then as you know, as the processor expands, and it gets more mature, then it might, you might move on to another model. So I know a lot of teams started off this way. For instance, at Microsoft, many of the groups like Office and Windows started off with the in house model I, I was part of it. At Facebook, it started off this way as well. So it's it's a good way to start off getting your toe into localization, and setting up all of the processes and and tools for the future so that you your model can accommodate the expansion that your company goes through. So having talked about the in house acquisition model, let me just open here the shot to see okay, I don't see anything that is really urgent. So I just max and the rest of the team from work from home will alert me if something comes up. So I'll just continue with the rest of the presentation. Thank you, Max. Appreciate it. So let's talk about the focalization model community. So if you are on the on the client side, you know, this might give you a little bit of a headache, because I'm sure you've been in discussions where people are said, Oh, why don't we use community, let's not spend any money doing localization, we'll just put it out there, you know, we will have people that out of the goodness of their hearts are gonna just come in, and do all of the localization for us. And you know, everything is going to be perfect. Unfortunately, it's not that easy. But let's go through the pros and cons of this model. So one advantage, obviously, is, is the cost. So if you have volunteers that are coming in and doing this translation, for free, you know, you don't have to spend that. So it's a big advantage. And it's kind of the first thing that people think about when it comes to mind, however, I think, almost are even more important is what is the engagement that you're gonna get with your current or prospective client. And I think this is something that you really should focus on, if you have a product that you know, you have a large community that is interested in, in it, and they want to be partners with, you know, they're advocates for you, they're participating in forums, they're eager to see what your company is doing, I think community might be a really good model, because then you can harness all of that energy and you know, get more from the get more input and get more engagement from that community. So having that customer connection should really be a top of the list you you shouldn't think about community as just like, oh, this is a way that I'm going to kind of get costs and outsource work that we should do, you should really think about it, how am I going to engage with my customers? How am I going to get more information from them? How can I use their feedback? How can I you know, give them more information about what we are planning, see how it matches with their expectation. So I really think this is the right frame to look at this also something that is good for community. And I mentioned this before that, you know, these are not like silos, these localization models, you can kind of pick and choose even within the same project to use one model for a couple of languages, another model for another language. So this might be a solution, where you are having either internally or externally demand to go into a certain market, but you are not sure on the ROI. So the return on investment. So you're thinking though, should I really be spending, you know, the, the money to go into, let's pick a market, Portugal when I already do localization for Brazil, and you know that there are reasons to do it other than the return on environment, but it might be a way for you to go to your clients and say, if you guys want to have this for your market, help us out and then see what is the level of engagement. If you don't have any pickup, then you might say we looked into it, it didn't work out. But if you have a lot of people that are eager to have this for their market, you know that you have a solution that is ready for that specific market. So it's something that you can use when you are thinking about expanding into new markets. So what are the drawbacks of this model? You have a lot of issues with the recruiting. So unless you are working on something that people are really eager to use, let's say for instance, game, let's say if you're working on Pokemon GO and you want to, you know people are almost addicted to the game and they want to have it in their own language or they are already using the English version and they want their family and friends to use a localized version or if you're working on content, where there's already an altruistic reason behind it. So let's say Kiva, which is a solution for lending money to the developing countries, and it might be hard for you to find people that are willing to dedicate their time and energy to helping out with localization. So it's, it's something that you have to consider well, and also that you might not have as many people available for some markets. So let's say you might have a lot of people that speak Spanish that would be available to, to do this work. But if you are talking about, let's say, Italian, you might not have as, as many people, so think about it, and how you're going to deal with this, and what are kind of the incentives and how you're going to get people to stay engaged and not burn out. So it's really hard to do the recruiting aspect, especially if you are coming out of a centralized office, if you don't have a presence in, in country where you know, people understand the culture, they understand the channels, it might be really hard to, to get people to work with you, then you have the question of volume, how are you going to do deal with volume, you might have people that, you know, they maybe give you 1000 words per week, but now your volume is at 5000 words, and you know, some of the languages have zero volunteer. So how are you going to deal with that? And how are you going to build a schedule, if you have no idea, you know, when translation is going to be that, so it's really tricky to deal with this, because you don't have a lot of leverage other than motivating people to, to do work. The scalability for volume, apart from having kind of your initial volume, as you get more and more into the development of your product, how can you scale for for that, it's really hard to, to do these things? How are you going to assess quality, if you have volunteers, and they're really passionate about it, especially if you are have a language that is covering? Many countries, let's say French, you know, you might have people from Canada, and people from France and Belgium and Switzerland and Luxembourg, you know, kind of discussing what is the best term? How are you going to assess that? How are you going to make sure that, you know, certain region that maybe has a younger profile is is a style that is much more informal. But you also want to go into a market where people care a lot more about traditional language quality. So it's, it's really hard to balance those seams. And then you also have to be very careful that people take the feedback in in the right way. If you're talking about professional translators, you may, you know, that they might not agree with you, but they will be professional and move on. If you're talking about volunteers, you know, they they might take it emotionally and say, Oh, you don't like the work that I'm doing. So I'm going to move away from the platform, and I'm not going to be engaged anymore. So it's it's really tricky to deal with those issues, and do the the arbitrage of what is the right term? What is the right translation? How can I kind of coach the volunteers, if their level is not the one that is needed, but they are very engaged in? So are they kind of creating more issues? Because they're just putting in translations that need a lot of rework? Or is there and some training that we can do and harness that energy to to deal with more volume? So it's, it's it's a really tricky model to manage. So when should you be looking into this? If you're working in the nonprofit sector, I would say this is probably the first thing that you need to look at, because you probably already have the right audience for it. You probably have people that are willing to help that they're eager to help. So think about that. Also content that is not urgent. So if you You have something that you're going to release next month, and you have to make sure that it gets translated by a certain date that you are going to, you know, have to go through final verifications testing, and all of that, and you have no way to control when your localization is done. The community is probably not that it's not the best model. But if you have something where, you know, it's documentation that, so let's say, for instance, TED talks are some sort of Academy, it might be something that, you know, maybe as even people go through the lessons, they will also translate them and get more quality into the translations as well. So something where you don't have a strict release date, is probably a better model. It's also going into markets where you don't have a return on investment. So people are asking you to go into that market, you don't know if you have the funds, or if you're gonna have the, the return. So you open it up to the community. And that's something that for instance, the Microsoft did a lot and people are asking for going into a certain market, either for Windows or for office, this would be open to the community, and a lot of times working with local partners, either on the government sector or nonprofits, and then they would try to get the volunteers and do all of the language quality items. And then you know, Microsoft, after we got the translations, we would release that localized version. So it's, it's a really good model. Funnily enough, it might be the one where you have to do more work and more engagement, and make sure that you go into it for the right reasons when when you go into it, because you just say, Oh, I don't have enough budget, I'm just going to ask people to help, it might not be the right model to go to, I think you have to go, we did with the very clear eyes and for the, for the right reasons for it to be successful. So let's talk about the next one. So the multi language vendor. And so this is the one that most people are more familiar with the one that they they use more. And so let's kind of go through the pros and cons and when when to use it. So what is the biggest fraud? It's that? Yes.
Max Morkovkin 32:56
Yeah, just want to tell you that we have already seven questions. And I think we need at least like 1012 minutes. So if you can get up a little bit, that would be great. Okay, five minutes be enough?
Jorge Russo dos Santos 33:08
Yes, I think 1010 minutes. So. So I'll, I'll finish in 17 minutes. And then we can have 10 minutes for question. You finished in in seven minutes, I need to finish in seven minutes. Okay. I'll speed up. So sorry, everyone, I'll try to move faster so that I can get to your questions, which I think it's going to be more important. So the pros of the localization model with MLB, it's that it's flexible, it's scalable, you know, you're talking with companies that do this for a lot of clients. So it's very reliable, you can ramp up the volume, you know, you have the processes and tools that are being used a lot. So you can have a lot of confidence in that. Some of the cons is the cost and control and dependency. So you're kind of putting your eggs into the basket of the multi language vendors. So you have to be careful about what you are doing. And it's a really good option for large organizations where you have a lot of volume, you have a lot of established processes, just make sure that you are not kind of tied into the process with with the MLB make sure that you know you still keep the translation memories that you keep your process agnostic so that if you need to switch vendors that all of a sudden all of your processes broken because we're using all of the tools from from the vendor. So keeping that your options open, I think it's going to be a really good option. So the SLV what are the pros it Still quality and specialization that you have of having a vendor that is specialized just in one languages. The The cons of this is that internal overhead and the communication is going to be much harder because in setting having a single point of contact, on the vendor side, you have many vendors. And you know, if you if you've ever met MLB, and they find an issue with one language they communicate internally, and they try to figure it out, if you have SLV is you have to be kind of the bottleneck in communication, doing all of the communication across all vendors, and you're going to have different levels of expertise. And so apart from the communication, finding these vendors, is really hard because you have to go in country and find the right partners. And you might have issues with the volumes and scalability. So as you increase volume, you know, you might be dealing with a company that doesn't have enough capacity. And if you want to go into new markets, you basically have to do all of the work again to find a good partner. So if your quality is a concern, like you really want to make sure that you have a really deep connection with the local market SLB might be a good option. And if your content is really specialized, and if you want to find a niche partner where you you work with them very closely. So there's also a model where you can have a team of freelancers, I didn't call it out specifically because this is kind of a hybrid of in house and single language vendor. And I think you can have such different experiences here. Because you might be working with a team of partners that you know very well, or you can just be going to a website where you have freelancers, and the first person that bids or has the the lowest bid works on your on your product. So I don't think there's this consistency where you can reliably talk about how to work on this. So I would say, with Question Mark said some of the pros would be the quality, specialization and sourcing. But depending on how you do it, this might not happen at all. And the cons are the internal overhead and the volume and scalability because you know, depending on how you do it, you can deal with these things very easily. Or if you are working with a small team of professionals, you might run into some of the issues that are similar to in house. And communication and process are also issues. So it might be a good option, if you have an organization that is very cost conscious. And if you have good internal support and processes to make sure that you give access to the team to the information that they they need. So workday as as an example, we use something similar to this, but we we use the freelancers that were dedicating some of their time, every week just to work, they and they were really experts, and we had a really high ramp up of up anyone that would start working in our platform. So what to look for in a localization vendor, make sure that you have an alignment with the video needs, current and future that you have subject matter expertise, you know, it might just be generic content, but you might be dealing with something that is very specialized games, aerospace industry, biometrics, any of those matters, you want to make sure that you have those subject matter expertise, that the technology and skills match, you know, if you are moving into our if you have a cloud solution, you want to make sure that people are familiar with that. And that there is a potential for matching growth. So if you are going into let's say, Middle East, you want to make sure that you have a vendor that already has a presence there. You don't want to be switching vendors just for that. What are the things to avoid, but this is something that comes up with bigger organizations like Microsoft, or Amazon where you are part of a small team inside a big client and you're not going to get you know, a lot of attention. So if your business is natural, if your business is not a priority for that vendor, trying to find someone that is more aligned, where your plans are in seeing where the markets that you wanted oh two, that company already has, as as a presence there, and don't get tied into tools or processes, try to keep it independent so that you are not. So tied in that later on, you can change vendors if you need to. So when use multiple vendors, I would suggest, do not split by language. Because this can create issues in your project with a lot more overhead. But you can do it obviously. And you can, if you're using SLV is this is something that you can do, I think it's totally fine if you split by projects, because then you can compare and contrast and you keep the issues just with one within one project. And splitting by content, I would advise you not to as well, unless you're dealing with something that is really specific, and you don't have the skills in one vendor. So I'd say subtitles, things along that nature, because you have to make sure that all of the glossary all of the translation memories are kept in sync and that people are sharing their information. So you can do it, but be mindful of the issues that you have there. So I think I went one minute, overtime. So I'm really looking forward to your questions and to getting to as many of them as possible. Yeah,
Max Morkovkin 41:32
eight questions so far. And let's announce the book. So George prepared humans by I think it's Brandon Stanton, for giveaway. So basically, we have I think, one minute per question. Oh, it's already done. Let's start one of the most popular by constantly and ranch. George, some companies are replacing human translations for some types of content with train machine translation. But there is the risk of embarrassing mistakes do you think communities can be instrumental in in spotting these mistakes before the content hits the Wide Internet? If yes, how would you set up mt plus crowd workflow potential this can increase the language reach exponentially.
Jorge Russo dos Santos 42:19
This is a great question. Thanks, Constantine, I, I would say it's actually the the, you also have to look out for the the other side, because, you know, I, I don't want to reveal details. But there was a company that kind of had egg all over their face, because their community in a certain market, they decided that it would be funny to do translations that used offensive terms, and then have some of the moderators, you know, they were all communicating the background, putting out offensive language, and then having the moderators going in going in and approving that. So there was no way for the company to know that that was wrong, because it was being suggested by the community. And then it was being reviewed and approved. So it took them a while to to get to that. So there's always ways that people can can game the system. But to answer directly to your question, I don't think we have solutions for generic machine translation to be presented to the to the public, right? Now you have to do a lot of pre work, you have to kind of restrict the corpus you, you have to write the content in a certain way to get more matches. So I think if you want high quality content, you want to have it human edited. So using the community to to do that would be one of the solutions. And I like how you are putting it but I think you also have to have some sort of professional validation because you might get an output that is so bad that even with community, if they don't have access to the source, it might be too hard to edit. So yes, I think it's it's a good solution.
Max Morkovkin 44:31
Okay, let's move to the next one. And we already have 1111 unanswered questions. So maybe we can post them on LinkedIn and you can keep keep answering. So then I choose the winner as an option. Okay, let's try. Mario Chavez USA in software companies when alone localiser has almost no visibility, how can he she reached to the top Doc's upper management to get more details about their localization plans?
Jorge Russo dos Santos 45:05
It's a great question, I think a lot of us have to deal with this, I would say just keep trying. I would say, you know, there isn't a big solution there. And even, you know, people like myself that have worked in the client, sometimes we have a really hard time either because people are not thinking about localization, or they are in a part of the organization where you, you can reach out to them, I would say, always assume the best and prepare for the worst, because in some cases, you know, people are going through reorders or people are working so much that they can give you information. But the the approach that I gave is, you know, I tried to be gentle in the in the first time that I interact with people. And if I get to a point where I don't have the information that I need, I will just say, these are the consequences. If you don't give me the information that I need, we're going to have to slip, the release date or the quality is going to suffer or these bugs are not going to be fixed. So you you put the onus back on the people that need to give you have access to the information, but it's a really tough one to address I agree with you.
Max Morkovkin 46:36
Or you will simply leave and go to another big company.
Jorge Russo dos Santos 46:41
That can happen as well.
Max Morkovkin 46:43
Another question from Martin crushtec. When do you think it's a good idea to leave the in house startup localization model? Do you feel that that would come naturally as localization needs expand?
Jorge Russo dos Santos 46:57
Yes, I think that that's something that comes up when you start reaching a point where you are going to so many markets that your team is going to balloon inside and it's going to be like a really big part of the the organization or you're starting getting into so much volume that you are already outsourcing the part of it as well. And so a lot of the times what happens is that like that initial team, they move more into kind of project management and engineering roles. And then you start outsourcing the localization, but you can keep those teams can with a 14 language quality to kind of make sure that things are looking good. So yeah, it's it's kind of a natural progression, because it just too costly to keep growing the in house team.
Max Morkovkin 47:51
A question from cloudy? Are you considering working with freelance translators at Amazon?
Jorge Russo dos Santos 47:58
I am not sure I'm the best person to answer that. But as far as I know, there isn't an option. Direct, that there might be options with one particular team where you might be able to do that. But I would assume now that it's it's really tricky for for people to work directly with, with Amazon.
Max Morkovkin 48:31
A question from Alibaba shave, what can you do to make transition from one model to another, less painful, we started off with the in house model and move to a combination of the in house and MLB models later on. And the transition was not as smooth as we hoped it would be.
Jorge Russo dos Santos 48:48
I would say, you know, having a really good process helps having reliable tools, where even if you have an in house team, people can work remotely and access things, having a lot of quality gates so that, you know, if you are moving to an MLP, you're not relying on having the same level of expertise as your in house team in terms of what they can touch and when and where they put the files and things like that. So I would say shore up your processes as much as possible. And then the transition to MLB would be easier. So, you know, maybe it's not the correct term, but kind of dummy proof, everything, look at all the past issues. Make sure that you have a process that those don't occur anymore, and then slowly transition to the MLP kind of hand holding and then as you feel more comfortable with with the MLP kind of relax a little bit. But yes, it's it's going to there's going to be some growing pains there.
Max Morkovkin 50:06
George, do you feel you already have the winner for book giveaway? Because maybe I'm sorry for interrupting. Can
Kate Vostokova 50:13
we take the last one? Because it has seven upvotes? Please? Okay. Yes, yes. The last one. I didn't see this. Okay, I
Jorge Russo dos Santos 50:24
would give the book to the humans from New York. It's a really nice book and also a really nice Facebook and Instagram account, I would give it to the one that has more upvotes.
Max Morkovkin 50:34
Okay, the more up for certainly read it. In an localization model where an agency handles all the translations. Does the localization manager have a responsibility to advocate for translator rights and wellbeing? Do we have a role in uplifting translators to ensure they're paid fairly and treated? Well,
Jorge Russo dos Santos 50:52
I would I would agree with with you. And I think that's something that we are all responsible for, on the client side, on the on the provider side, and I, I tried to advocate for translators as much as possible, for instance, in terms of giving them access to information and context and because it's, it's, it's kind of a common mistake for people to say, oh, localization quality is not good, because, you know, people don't have the best skills. And I always say, if you don't give people the, the right context, they want people to do a good job. So rather than kind of putting the onus on the translators, or, or the vendors, I always go back to saying, oh, let's give people a chance to be successful. And so that, that goes with the time giving enough time for people to do the good work that goes with cost, you know, if you're trying to cut corners, everywhere, your quality is going to suffer. And that goes with the context giving enough information for people to do a good work. So I wholeheartedly agree that we need to make sure that we give our partners so working on a client that we give our partners chances to give their employees a really good role and a sustainable salary and all of that.
Max Morkovkin 52:31
Okay, then the book goes to be our TR e one. Are you the son or daughter of Elon Musk? very strangely nickname. Okay, so we will share the book with this person. Thank you for asking the questions. George, thank you for a very nice presentation. You see, it's really interesting, so many questions we had today. We appreciate your presentation.
Jorge Russo dos Santos 52:57
Thank you, Maxine. Thanks. Thanks, everyone. And if you have any more questions, just put them on LinkedIn or contact me and I'll try to answer as many as possible.
