Andrew Thomas |
Welcome to Globally Speaking. My name is Andrew Thomas, senior director of marketing at RWS. I’m joined here today by two senior managers from Dell technologies, Fernando Caro, and Scott Cahoon. Would you guys mind introducing yourselves quickly for our audience? |
Fernando Caro |
Sure. My name is Fernando Caro. I am a senior manager responsible for internationalization at Dell. I have a group of nine folks who work to make Dell software global in nature, and also to make it easy to translate without requiring a lot of engineering efforts. |
Scott Cahoon |
All right. And then I’m Scott Cahoon. I act as the internationalization architect for the translation technology play at Dell, and so work closely with Fernando and the rest of our internal translation team to make sure that all of their needs are met as far as translation tech is considered. |
Andrew Thomas |
Perfect. I’m really excited to have both of you on the podcast today, because I think very often when I go out and listen to localization presentations or podcasts similar to this, it’s always either from a services client side or delivery perspective, or it’s very much a technology driven conversation. I think we rarely take a step back and look at how the two intersect and obviously they very much do. Clearly they do here at Dell. I’d love being able to present the entire picture for everybody. I do want to start there at that higher level. And maybe I’ll direct this question first to you, Fernando. Talk about what are the larger business goals at Dell that are directly supported by localization? Where do you see your role in how you contribute to Dell’s business? |
Fernando Caro |
I think to be one of the most major bigger roles is as the world shifts, not just in terms of what it’s been doing for a while now, in terms of agile, continuous integration, continuous development, is the as a service world. In that world, you really can’t have delays or make your customers wait to get whatever service, subscription, product that the as a service world is offering. And so I think from that angle alone, it’s important that, well, both from an internationalization, but from a localization continuous look, that those pieces be rightly integrated into how that entire set of workflows work to enable us to deliver simultaneously the experience from my local experience at the same time as English is going out. |
Fernando Caro |
I think that itself is a major driver of being part of that customer experience, that customer journey to enable that to be a reality, but also enable the languages to be part of that experience. So that customers get a seamless end to end journey while they traverse the, as a service world. I don’t think that today that’s an inevitable future, it’s here now. |
Andrew Thomas |
I think that’s a really good point. I completely agree with you. I think across the variety of industries, we’re seeing this shift to the, as a service approach as you call it. Clearly even if you’re not making that shift as a business, as you mentioned, customer experience is really important regardless of the nature of your business. Would you mind just unpacking a little bit, the sorts of issues or challenges that you face when it comes to localizing Dell’s customer experience and looking at it from that point of view? |
Fernando Caro |
Sure. One of them, first of all, is just understanding what is the customer experience. If you get right into simply the translation part of it, you probably are not looking at the world with all the right sets of lenses. What I mean by that is, you have to understand what is the customer actually doing with that as a service world? What are the learning experiences, the buying experiences, the end to end throughput to essentially then understand, well, what would be that language experience? Are they getting bounced around from locale to locale? Are they kept within their user location? And sometimes those things get ignored because you’re thinking purely about what we need to translate into X number of languages. I think it’s not so much about which languages or the actual translation process, it’s really about mapping how the customer will traverse that as a service world and what will they get and how will that work when you’re not in an English speaking country or region or domain. |
Fernando Caro |
I think that’s the piece that you have to really first analyze in order to understand what needs to happen, because behind the, as a service, of course are a set of services, products. You’ll have to look at that side from understanding, well, what is supported? What is not supported? And even minor details like translating a user interface as part of the, as a service world, but if the products themselves do not have the capacity to handle data that’s not ASCII, then you don’t have a great customer experience. You might have user interfaces in their local languages, but how they operate within that as a service world, won’t give them that end to end experience. That’s what I mean by trying to really delve into what is that journey? What are they doing in that journey to then map correctly what that experience will be when you go beyond English? |
Andrew Thomas |
It sounds a lot like we’re taking the older internationalization approach that we’ve always applied to product development for a long time and now extending it to what has become the larger product experience or service experience, to your point, which is much bigger than the confines of an individual product. I imagine you have to go a lot further upstream than you would have in the old days, being in a loc department. I imagine you’re having to work very closely with other teams much further, like I said, much further upstream in the development of these things. Can you give us some details or some examples of that? If I’m correct. |
Fernando Caro |
I think the key here is, is that, at the end of the day we all know what to do with globalization and the processes behind it. But when you look at this world and you look at how companies are transforming their data centers and how they operate, you can’t really look at globalization as your traditional steps one through 10. You’ve got to take a step back and realize that the world that we used to live, which is you engage with a product team, you do what you have to do to globalize it, translate it, that’s still true. That’s not going away, but the, as a service world is a different paradigm. Because you’re having customers interact through essentially a portal, but they’re interacting through this different mechanisms. They’re not buying a product. |
Fernando Caro |
And so I think the examples for me are more around, and this is where it’s challenging, because it does take you to places that you perhaps don’t traditionally look from a localization or even globalization perspective, is what’s the journey the customer takes? Where are they going through the, as a service world? What are the expectations that they’ll have based on what is being provided to them in terms of selections, not options? You have to understand obviously the competitive landscape, you have to look at what is the actual support that each of these areas that comprise the, as a service, what do they support? And that’s what I was talking a couple of minutes ago, which is, it’s not just what you translate into, it’s what functionality exists as they purchase, go through the, as a service, what is the experience you’re giving that customer and is a customer then going to have an ability to purchase because they had a seamless experience in their native languages, but in their native functionality as well? |
Fernando Caro |
It does the customer no good to have a translated interface if the underlying technology doesn’t understand when the data is not uniformly ASCII, and when they have to do things that are specific to their region, in terms of formats, things of that nature, if that’s not there, then they get a half baked experience, because what they’re getting is, yes, native experience in the UI, but then how they operate isn’t really, truly matching what they would expect from their country or from their region. |
Andrew Thomas |
Right. That makes perfect sense. I do want to bring Scott into the conversation now, just to talk about the translation platform at Dell. Just taking a step back, obviously there’s a lot that Fernando’s doing to support Dell’s localization efforts and you are providing the technology platform that his team relies on. Can you just give us a quick background on some of the capabilities that you ensure that you can deliver to his team, the history of usage of technology at Dell, any future plans where you’re taking things from a technology perspective and how that integrates with his team? |
Scott Cahoon |
Yeah, sure. I think one of the things that we found, as we are moving more towards the SAS model, and as we’re looking to see how we can really have a larger impact on the overall Dell ecosystem, is that, our other parts of our business are beginning to, I think translation is always being considered something you have to do to do business in certain places and maybe to be successful with certain customers. But what we’re seeing happening at Dell, is that some of our core teams are beginning to see translation as an enabler to them to run a more efficient business. A good example of that would be, our support team has recently been looking to see how they can better support our international customers and really there’s two ways you go about doing that. One is you spend a lot of money, you hire a lot of resources and you locate those resources all around the world. Right? |
Scott Cahoon |
Another option, which we’re in the process of putting in place right now, is to be able to take technology and take advantage of the latest neural machine translation solutions and look to build a real time translation solution that allows your English support agent to now interact with the Chinese customer, and to be able to successfully walk them through their issue and provide a resolution without necessarily speaking Chinese. Right? We see those kinds of asks coming to us. I think in general, the neural MT or machine translation as most people outside of our industry would understand it, is beginning to open a lot of different opportunities inside of the Dell ecosystem. |
Scott Cahoon |
And then I think you take what the neural machine translation can offer, plus then looking to see, as Fernando stated, as we get to more of a SAS, always online, always available solution. What it’s starting to do is force our development teams to get out of their silos, right? You can no longer say I’m just an English, I’m an American only product, therefore I don’t need to localize and I don’t need to translate, and I don’t need to internationalize. What’s happening is, those development teams at Dell who have for years resisted or not felt the need to become international are now realizing that if they don’t become internationalized, if they don’t properly support our international customers, they’re going to basically fall by the wayside. Right? |
Andrew Thomas |
Because they’ll introduce those problems that Fernando was talking about. |
Scott Cahoon |
Exactly. |
Andrew Thomas |
The underlying pieces and is enabled to support the local experience. Yeah. That makes total sense. |
Scott Cahoon |
Exactly. So as somebody goes through the SAS experience, we don’t want a customer to start out in a nice, fully internationalized, very fully translated solution and then suddenly click on a link and pop to an English product, right? We want them to have a consistent experience as they go throughout the entire solution, and that requires that we move to a standard set of languages, that we move to a standard level of support. And that’s where I think, Fernando and his team are really getting an opportunity to shine at Dell, because the value of what they offer is really becoming center stage. |
Andrew Thomas |
Let’s flip it around a little bit. Fernando, how do you see Scott’s platform supporting your team? What does Scott’s team and his technology bring to the table for your process? |
Fernando Caro |
Well, first of all, stating the obvious, it’s extremely critical to the entire equation, while for the business those types of areas may not be visible to them, but in order to make this all work, it’s not just what I was talking about earlier, but it’s also the technology pieces, because you can’t operate in a heavily transactional manual way. While this part doesn’t typically get surfaced or visibility when you’re talking to the business owners, the folks that are making the SAS world possible, but we need these pieces, because all this revolves ultimately, whether it’s agile or other models, you need that continuous integration, continuous development. And the set of technologies behind that are critical to remove manual processes that would otherwise be impossible to work with in terms of this model. I see the entire set of stack of technologies that are behind the scenes as a pivotal foundation to making this all happen. |
Andrew Thomas |
Well, yeah. Speaking of that, you mentioned continuous development and a little bit earlier, you mentioned continuous localization, I’d like to unpack that a little bit here since I think a lot of people have different thoughts on what continuous localization is. Would you mind explaining what Dell sees or what Dell’s continuous localization process looks like? How do you define it? Why is it important? What are the, maybe some benefits and some challenges that you’ve experienced in moving to that process? |
Fernando Caro |
First, I see continuous localization as entirety of globalization. For us, we define that as internationalization and localization. It’s not just what you have to do to the transition workflow process to enable that to happen, you have to first start with what you’re doing to make that code, that software application, that product, service, work globally, and that needs to happen. That requires a lot, because it requires a cultural mind shift change for most engineering organizations who do not see this as either a core to what they do, and or, an important aspect of what they need to do. You have to put those pieces in place and those pieces are not, obviously there are things like freeze dates, you put rigor behind governance, but you need to build that cultural mindset around, that spans by the way, the entire ecosystem to what you need to achieve. |
Fernando Caro |
You need that piece in place and you need to work at it in order to then make the continuous localization process a success. That’s where I was inferring a little bit about where technology comes in, because you need technology on the internationalization side to make things efficient. And really, our purpose to do this at Dell is to integrate internationalization within all the existing engineering workflows, to enable engineering and I mean development and test, to internationalize code and tested in line with how they operate today. It doesn’t work if you take them outside their systems, it’s a very hard sell if you’re going to do that, especially in today’s world of limited resources, time constraints, pressures to deliver, deliver, deliver quickly. You really have to establish that process through, not just a set of steps, but through technology. |
Fernando Caro |
You need to infuse things that will automate the software localization workflows, how development can quickly assess the state of their compliancy with internationalization through static tools. That’s a big component. If you dive right into continuous loc thinking that it’s a localization process, you probably won’t get there. If you do, you get there through a very painful journey. Then I think then, yes, you need to look at the localization process. You need to also have the right sets of tools. You need your LSPs to be in line. You need to obviously understand how you have to adopt the traditional way of translating into a model that doesn’t have capacity for delays. But to me, you really need that front piece first in order for the whole ecosystem to work. |
Fernando Caro |
For Dell, it’s critically important. The days of having to think about plus 30, plus 45 and further out are not a thing of the past, but it’s really not in line with how the world is moving. It’s not how our customers operate. It’s not how the business operates. Scott touched on it a bit, but also from a revenue perspective you want to allow the business to operate in that manner. I think continuous loc plays a huge part of driving the business and making it possible for regardless of where you’re located around the world, to get whatever you’re going to buy, whether it’s through the SAS model or otherwise, buy it when it’s ready and not have to have a delay in how you purchase things, because you don’t deliver on a continuous loc. Bottom line is I think continuous loc is another critical piece in the equation. |
Andrew Thomas |
Yeah, I couldn’t agree more. I see the trend of, again, I see this across different industries, even when you get to step outside of the high-tech industry where Dell sits, and you look at other industries, they’re all shifting towards embedded software capabilities with their products, which means suddenly everybody has to consider high-tech, consider needs. The world is just, markets are expecting products simultaneously everywhere. And if you don’t ship everywhere all at the same time, then you wind up losing to competitors or knockoffs. Right? And so it does seem it’s just becoming the new norm. Certainly this last year with everybody being stuck at home and having to rely so much on technology and working distantly from each other, has I think increased that adoption or that expectation that they’re going to get access to things as soon as they need them, rather than the traditional the older tiered approaches. |
Fernando Caro |
And if I may, Andrew, there’s another benefit and it’s not often thought through correctly, and it’s not, is obviously describing a lot of the external aspects to the equation. But I think internally it’s also critical. Engineering from both a technology or technical perspective, but also from the business, they can’t afford to divert resources to create additional layers. I touched upon it a little bit when I was talking before this. It’s really a hard sell for them to do something when they don’t have the time and you’re asking them to do something that doesn’t fit their model. It doesn’t fit their business operations, how they operate, and of course they have a lot of pressures to deliver on features and deliver on things that customers are asking for. |
Fernando Caro |
One of the internal benefits of a continuous loc world, is that you, as I described it earlier, you enable them to do what needs to happen, because there are business requirements. We have to translate into a certain set of languages, and that will vary of course, in what you’re talking from a product perspective. That internal benefit is, is that, for them they don’t have to divert their engineering time and resources to accomplish something that takes them away from other important aspects of their business. And obviously that primarily comes down to features that are being requested by customers and so forth. I think there’s also an internal benefit at this model which is not obviously customer facing in the sense of what you have to do, but it does impact the whole ecosystem. |
Andrew Thomas |
Yeah, that makes really good sense. We talked about some of the benefits, but I know that this is a challenging shift for most companies. It’s certainly sometimes a challenging shift for LSPs in the way that they interact with clients. What are some of the downsides? What are some of the challenges that you’ve faced or that you see just in general, when people move to this process? |
Fernando Caro |
Well, we’re still at it, the journey is never really complete because you’re always having to adjust and to see the world not as a uniform way of thinking or doing. The challenges are to some extent, pretty much what I was talking about, cultural mindset. People have to think different differently. They have to approach things not in a transactional serialized way of executing, but where you have to align processes into a parallel world. Some of the challenges that I think and I would think that this is more of a global thing, is built around changing the way people operate and the way they think, and that’s not as easy as saying it here, it takes time. It takes time to adapt to what needs to happen. I think another challenge that stretches is, really understanding who your stakeholders are, what do they really need? Why do they need it? |
Fernando Caro |
It’s not just, okay, we have to set up an automated workflow on the localization side and the LSPs as well are art of that picture, it’s understanding what is engineering really asking for? What are the executives looking for? Why does that matter? Because that impacts what you have to go do. A challenge is, is understanding your stakeholder from a different angle, because you’re not asking them about what volumes you have. You’re trying to really infuse yourself. That can be a scary world, because if you’re not from an engineering background, it can put thoughts into people’s minds about how you engage an engineering organization to further understand how they operate. Those are subtleties. I know that obviously there’s challenges with technology, having the right setups, making that ecosystem work, turnaround times. |
Fernando Caro |
I think those are the more obvious ones, but I wanted to call out those because I think those are the ones that typically if you don’t think about them first, you’ll face them, but you’ll be facing them without really having sort of, you have to face them on the fly and that causes a lot of churn. For me, at least I think those are two of the big ones that I see beyond the traditional way of thinking around the tools and what we do today from a globalization perspective. |
Andrew Thomas |
I’m assuming when you say that you have to socialize this concept, you’re talking about it across the board with stakeholders, right? Because you have to get engineering to think a new way, your own team to think a new way and the LSPs to think a new way. Right? |
Fernando Caro |
Yes, it’s the entire ecosystem. You can’t really look at it. That’s why I was talking, when I was saying about different sets of lenses or angles, yes, it’s the entire ecosystem because you can’t just look at it from the approach. I think the first one is to really look at your stakeholders. It’s easy to go where your comfort zones are with what and what you are an expert at. But I think the first thing is to really understand what are those business requirements? Why does this matter? Because those things, why does it matter? At the end of the day, I know I need to do continuous localization. I need to set up the workflows. I need to do this. I need to do that. But those things can be very impactful, because those things can derive or at least point you in different directions and either take you in the direction you need to go, or it could take you off your direction. |
Fernando Caro |
I think the challenge there is that you have to think, and it’s a cliche to say this, but you have to think outside the box. You have to challenge your own mindset to not just think about the term continuous localization and how do you achieve simultaneous translations, if you will. I think for me, that’s where I see, the big area of thought that needs to be happening before you really get into the backend specifics. |
Andrew Thomas |
Well, I do want to talk about the backend specifics a little bit with Scott. Fernando was mentioning how you guys had to be more tightly integrated with the development process, the build process, the engineering process. I assume some of that is down to the hooks and the ways that you integrate the translation platform that you have at Dell in with those processes. Can you talk a little bit about that and what your experience has been in supporting this approach? |
Scott Cahoon |
Sure. One of the things that we have seen at Dell over the last, I would say probably year, year and a half, is a huge increase in the desire for our development teams to plug directly into the translation solution. And so that goes from our websites, our web apps, all the way back to our applications. And as we move more towards our SAS solutions as well, they’re looking really for the opportunity to really, in some ways, there’s this challenge of how do we redefine the way they develop even, right? Because if you think about standard software development processes, one of the challenges has always been who writes the interface, right? Who edits it, who controls it, who manages that, who makes sure we’ve got high quality content there. |
Scott Cahoon |
What we’re finding is that we moved to the SAS model where it’s almost a mix between marketing and software, right? We’ve had challenges where now the authors of those interfaces want a better interface to work in, right? Traditional software doesn’t really provide that for you. The place you go to get that typically is our content management systems. And so we have a number of content management systems plugged into our solution. But one of the challenges then is, how do you mix the two models, right? You got a group who wants to author and content management, but what they’re building is a software solution. And so how do you then provide the flexibility to iterate the interface so that you get a clean interface that works properly for your customers while still providing a robust authoring environment for the creators of the English versions of the interfaces, right? |
Scott Cahoon |
And so that’s definitely a challenge we run into. I think one of the other challenges we run into as we move towards this SAS model, and also as we move more towards continuous translation, is that, in order to maximize what we get out of our technology platform, there’s also a need really to standardize, right? Our processes, because in order to automate and reduce overhead, which is what really the goal is here, right? How can we move from development to an English product, to a localized product, to release product, all in as tight a timeline as possible? The key to that really is how do we eliminate waste in process and how do we minimize is overhead in the translation and release process, right? And really that comes back to two competing, I would say elements of translation. |
Scott Cahoon |
One is the automation and optimization. How fast can I get something out versus how much time and effort can I spend in making sure it’s high quality, right? That’s one of the areas that we’re really looking at right now, which is, how do we bring quality? We don’t want to sacrifice quality for speed, but in order to optimize both, one of the things that we’ve historically done a lot at Dell, is to try and be a very, I would say, overly accommodating. We have allowed our stakeholders to get used to an idea of, for this job it’s okay to skip a step. And for that job, we can add a step. And for this one over here, we can do a little bit of a tweak and change. And so what that ends up doing is building a very complex environment where in 20% of their jobs, you include a client review and then the other 20% you don’t. |
Scott Cahoon |
In this case, we rely on a vendor to do that review. And in this case, we rely on internal resources. What we’re finding is that in order for us to really be an efficient, an effective and highly optimized solution, a lot of this goes back to, beyond the technology, it goes back to process and standardization of process, right? |
Andrew Thomas |
Yeah. It sounds like you’re basically saying something very similar to what Fernando was saying around, in the process of that standardization, having to go back and engage all of the stakeholders and get a mind shift if you will in their approach, given your past customized approach, if you will. And that’s in conflict with standardization. Right? That makes total sense. Although the topic of quality is actually very fascinating to me and I’d love to take a little side trip on that if you guys are willing. Because I find that so often, particularly in the localization industry, we talk a lot about quality, but quality for the business oftentimes means different things, right? If you’re talking within the loc industry, it’s traditionally linguistic quality and how well something is translated and representing what was the original intent of the original language. |
Andrew Thomas |
But going back to your earlier customer support example where you had moved to using NMT, the quality in that scenario isn’t really about the quality of the translation, it’s about the ability for a non native speaking support person to be able to resolve a problem that somebody who speaks a different language is having, it doesn’t really matter how good the translations are, presumably it doesn’t matter if they’re able to resolve their problem. Right? |
Scott Cahoon |
Yeah. When we talked about the application of MT in that scenario, what we’re looking for is the metric you want to measure against is not necessarily, was the translation perfect or not? Right. What you really want to look at is, did the translation enable completion of the task? Right. That’s one of the things as we’ve been moving into this, we’re in a POC of that solution at the moment, one of the things that we’re working hard to set expectations with our team members who will be involved with that, is that we shouldn’t be overly concerned about the dot the i’s and cross the t’s, was that the term I would have chosen to do this translation, right? What we want to focus more on is, did this give you the information to be able to understand the customer’s concern and therefore provide a resolution for them? |
Scott Cahoon |
When you provided that resolution, were they then able to understand the resolution and get their problem solved? If those two things can happen, then the solution is successful. Right? That’s something that I think is really interesting to try and attack is. And then also inside that whole support arena is there’s, as you probably know, with knowledge bases and other things out there, support generates a ton of content, right? It’s the one area inside of our ecosystem where we do today, support delivery of raw MT without any editing. That team is actually taking the lead on really pushing us to expand the coverage of raw MT and we’re in the process now of looking to move something languages fully to raw MT solution, and then see how the customers react to that and look to see their feedback and see whether or not it is enabling them to get the information that they need, to be able to get the resolution to their problems. |
Scott Cahoon |
What that does in the end, is it enables us to, one, provide more content in those languages, as well as expand it to support localized content in more languages. Right? It’s definitely pushing us and the MTs engines capabilities to see how well we can make that work and whether or not we can get the quality we need to be able to have a successful solution. I think in our general translation work, as we talk about quality, what we’re looking at is, how do we build quality processes that ensure that our customers get what they need and are able to get high quality products, because often quality and speed are your trade-offs, right? And so how do we minimize that trade-off while still maximizing turnaround time and also quality? |
Andrew Thomas |
Well, it’s funny you bring that up because I wanted to ask Fernando, the quality impacts or shifts in perception of quality or how you approach quality control on your team’s side in a continuous loc process, because I think to Scott’s point, speed sometimes is in direct conflict with quality, but also the potential advantage you have with continuous loc is that you can iterate the same content more than once. Whereas in the old days you may have only gotten one crack at it. I’m just curious what your approach is on your team when it comes to quality and how you think of quality in the work that you guys do. |
Fernando Caro |
Well for me when I think of quality and I think this is another one of the challenges back to the question you were asking a couple of minutes ago. In my view, right? In the globalization business or industry, everyone wants to have the best quality, right? You don’t want errors. You want, I wouldn’t say, perfection, but close to it. In a continuous loc world, in this world that we were talking earlier, the question is, do you accept imperfection? Because you’re having to basically align to a very rapid environment where there’s literally no delay in how things get developed, tested and ultimately released. And so the traditional model of how we’ve approached that aspect in globalization doesn’t quite align to that continuous loc, CI, CD world. I think the challenge there is, is it okay, is it acceptable to say I’m not going to shape what the expected quality output that we have defined, there are going to be errors, and is that acceptable? Is it acceptable to the company? Is it acceptable to our customers? |
Fernando Caro |
One can make the argument that in the non translation world, when they’re in this world, they’re shipping with bugs, they’re shipping with features that are not 100% yet complete, but that’s the whole point of this rapid development world, is you’re not in that old world of waterfall where it took you six to eight months to produce a beta until you had everything lined up as if it was ready to ship. In this world, you put things out there knowing that they’re not perfect, but you want to get that constant input from your customer base. And that’s what you iterate on and then you go and improve and change and then you work on it. If it’s okay from an engineering perspective to accept that that’s the reality, why would it not be okay from a localization perspective to accept that the quality isn’t going to go out to the same level of expectations in that world than in the traditional world, knowing that you ultimately will get to the quality you expect, but it’s not going to be in that same fashion. I think that’s a big challenge. |
Andrew Thomas |
Yeah. It’s almost like you have a new demarcation line. I know in engineering speakers usually talk about a minimum viable product. And so it’s like a minimum viable quality, is certainly like a concept maybe, or a way of thinking about it. That’s interesting. In my mind, I do see the iterative nature of continuous localization as being a potential for actually improving quality more so than the old processes, because in the old days you tried to get everything perfect, but being perfectly honest, mistakes still happen. There is still quality issues, even before we all moved to continuous localization. More often than not, when the quality errors happened back then, you had to do a big dot release or you had to wait for another major release vehicle in order to fix whatever major quality issues there were. |
Andrew Thomas |
I guess one of the advantages with continuous loc is, I assume you let me know if this is true at Dell, is that that’s no longer such a pain, you can improve the quality and attach it to the next iteration, whenever that might happen to be going out. |
Fernando Caro |
You can, but that’s part of that culture shift and that mindset shift to understand that in this world, it’s not going to be the same world that we all lived in when we were doing traditional globalization engagements, mostly with product engineering. And of course in today’s world, it’s not just traditional product engineering, you have a lot of other software development that takes place in different forms with different customers, different audiences. I think that’s part of that challenge that we were talking about earlier. That stretches you outside your comfort zone, because you know that, for example, a sentence is completely wrong and in that local language, well, what will it mean to the customer? Did you just create a geopolitical issue? Did you give them information that doesn’t lead them to know what to do next? Because they don’t understand what’s next in the journey. |
Fernando Caro |
I think it’s okay, if you think about it from that broader perspective that you’re shipping in small increments, you’re not going to be perfect out of the gate. You’ll get there, but each increment gets better. So yes, to your point, I think it does improve quality from that angle, but it’s acknowledging that that’s what you’re doing. Your mind doesn’t think that way. Right? And so your mind needs to change to see that this is actually a good thing. It’s just a different world that’s necessitated things to be looked at differently. For me, I think, yeah, you are potentially improving that quality, it’s just you’re taking a different route in doing so. |
Andrew Thomas |
Yup. That makes total sense. Just wrapping up the conversation now, thinking about, we’ve talked a lot about your focus on the customer experience, your shifts and adoption of continuous localization, the technology platform considerations, the quality considerations. Is there anything else around this that we haven’t already discussed from either of you that you think is a major element, that listeners to the podcast should be aware of when it comes to either adopting a continuous localization model or taking a different look at how they interact with product development teams or product experience, customer experience, those sorts of things? Is there any aspect that you think, any best practices, any major takeaways, any major advice that you can offer? I think our listeners would really appreciate that. |
Fernando Caro |
I would say for me that when you look at this world, this as a service world, how does globalization fit into that world? How do you make it work? I think you have to see it, not as a point in time journey, but a real journey. You have to be patient. It does take time, because there’s a lot of moving parts. There’s a lot of considerations that need to be taken, and you’re not going to get there in one piece. It’s similar to what we were just talking about with the quality aspects, you’re going to get there, but it’s going to take an element of time, a lot of perseverance, a lot of patience. But understanding that that journey, even if it’s a multi-year journey, will get you there. You’ve got to incorporate the bumps into the equation. I mean that pervasively from whether it’s the process, the technology, the combination. The goal isn’t to get there as quickly as possible. |
Fernando Caro |
In my mind, the goal is to get there in a way that allows you to ultimately achieve your entirety of your goals and your expected outcomes. And that’s why I say it’s more of a journey. That’s not saying that I advocate that it takes multi-year journey, but you can’t squeeze everything into a neatly defined set of steps. It’s going to take you time, and it’s also going to take you away. You will find out that that journey isn’t going to always go the way you expected it. I think that’s one of the takeaways that I’m learning as we go through our own journey here at Dell, about how to establish that model. Also keeping in mind that the model may not be uniformly the same if you’re implementing that model on different business operations. |
Fernando Caro |
Product engineering might have a continuous loc model but the dot com space might have a slightly different model, because the business and the requirements, the audience are not uniformly the same. You have to be, I guess, at the end of the day, adaptable, that’s not to say that you have to create any number of loc models that continue as loc models. You certainly have baselines, there’s definitely common denominators, but you have to be careful not to think that the world is uniformly the same and try to fit that model exactly the same way across every aspect of the company’s business operations. |
Andrew Thomas |
It’s back to that same struggle that Scott was talking about around standardization versus offering customized approaches. Scott, I wonder for you on the technology perspective, as you think about the history of translation technology at Dell, where you started, where you are today, where you’re headed, same kind of question, for our listeners, if you had to give one piece of advice or one best practice when it comes to building out a translation technology platform, whether or not they’re going to support continuous loc, what would your advice be for our listener? |
Scott Cahoon |
I would say when it comes to technology and translation, especially translation and technology, unfortunately in our industry, you’re never going find the silver bullet technology that solves all of your issues. And so I think what you have to do is you have to decide what you really want the platform to do for you. And then you need to look, don’t be afraid to, one, make mistakes. You can go in, as you build your solution it’s okay to iterate and to have trial and error. I think as well, I think one of the other key is you need to make sure you understand that, building a solid translation solution takes more than just process. It takes more than just technology. It takes a combination of the two working in a nice rhythm and in good synchronization, right? |
Scott Cahoon |
You can have the best technology in the world and if you have bad processes, it will do nothing for you, right? You can have the best processes in the world and if you have bad tech, it’s going to be hard to make those processes bring the profit that you want them to do for you. Right? And so the key is to work to build both out and to maximize both together. And if you do that, then I think if you can make your technology do what you need it to do, if you’ve got a good process that takes into account what your technology is capable of doing. That’s as locale as is today. I think as you look towards your future, then you can look and say, let’s to get to a steady state, a solution that meets our needs. And now we can look to see how do we add more bells and whistles to that solution. |
Scott Cahoon |
How do we find that next iteration, that next gen solution that maybe moves us closer to that nirvana of a translation platform that also provides all of the program management, project management, globalization management solutions that we’re looking for. |
Andrew Thomas |
Well, it sounds like a great summary for at least my perspective on the way that you guys interact with each other, talking about process and technology and working in lockstep with each other and iterating. As you say, Fernando, you guys are on a journey, but I think from my point of view, you guys are pretty far along that journey and it’s very impressive. I want to thank you both for sharing your story with us and letting us peek behind the curtain of what goes on at Dell in your localization processes. Thank you for sharing that with our Globally Speaking listeners. |
Fernando Caro |
Thank you very much. Appreciate the opportunity. |
Scott Cahoon |
Yeah. Thanks a lot. |
Andrew Thomas |
Thanks guys. |