← All events

A (much needed) paradigm shift

February 10, 5:39 PM
YouTube video player

LSPs are facing pressure on gross margin, top line growth challenges, and business model relevance. Only the larger players benefit from high volumes of business, the critical factor for profitability through economies of scale. Marc addresses the challenges of growth and profitability and provides principles and models to stay resilient in the midst of a health crisis and an increased efficiency of machine translation.

Transcription

Max Morkovkin 00:06 Hello, Mark. Hey. Marc Jonckers 00:09 Hello. Max Morkovkin 00:10 Can you hear me? Can you see me? I can see you. I can hear you. How are you? Marc Jonckers 00:15 Great. Great. Great. Thank you very much. Max Morkovkin 00:17 I just announced the book your way. Maybe you can tell a few words about that book to encourage? Marc Jonckers 00:25 Well, yeah, I'll speak about it in a minute, because it's part of the presentation. Max Morkovkin 00:29 Okay. Marc Jonckers 00:31 Shall I load my my presentation, by the way? Max Morkovkin 00:34 Yeah, you can start sharing your screen. And this presentation will be the first in our mindset track. So I'm very excited. Sounds like some brain work. Right? Yeah. Marc Jonckers 00:50 So I wonder whether you see me on my screen, do you see my screen? Max Morkovkin 00:55 We can see your screen we can see you. So okay, good. So I shall start. Marc Jonckers 01:00 Thank you very much. So a much needed paradigm shift. Indeed. I had a major paradigm shift in 1994. When I when I found that young crews were was called in by new concept at that time was called the Internet. So obviously, in the meantime, 25 years has passed and I sold my shares and started a new business. I'm now an advisor and a consultant for star and scale up LSPs and language technology companies. Change management is my passion. And it's obviously my my turf. So I'll share my perspective. with you today. It's based on my experience in my perception of the market, many of the topics tackled today would require a session for their own. So do bear with me if I'm, if I'm picking shortcuts. So, let's start. Kill your darlings. I'd like to recommend two books. So Max, you asked me. First one is Eric Ries the lean startup and the other is a very well famous of very well known and famous Daniel kin, Kin mins Thinking Fast and Slow. To me Combined, these two books, they form a great mental equation, cognitive biases, and you know the need to pivot when those biases go wrong. This is really the name of the game. So, obviously, changes are generated by crisis, internal crisis, external crisis and, and opportunities as well. So those changes can become pivotal change in service offering changes that have infrastructure and organization go to market and so on. Key is that private thing leads to change in the way organizations operate. That is fundamental. That change will induce a Kill Your Darlings attitude, that's means to me, abandoning how you like to do things, and how you like things to be done. Innovation, have a new perspective about your business and your market. It's a process, it's definitely not a one off, it's not a one time event. And at the core of that change lies the new value proposition you need to have matured, until you have material evidence that your customer traction is optimal. And not until you believe that you're doing things better. What you think doesn't matter. In the end of the day, what matters is what's your customers experience with your service offering or your product. So when you carry out momentum of innovation, it's transformative. That impacts your business model. And your business model has a big impact on the way you generate profit. So usually, it means bringing fresh blood in your organization and understanding that it is important to challenge your own ideas, and have them challenged by people who might think differently. So technology. Now that's a buzzword, but in our industry, I mean definitely we all tech driven. But digital transformation, which is a big word for I would say organizations that are relying on heavy industry processes. It's important for them but for us it's like it's like a second nature. We've always been driven by technology. Now one important question is whether we should develop our organization or create our organization as tech based like like software development organization, for instance, or like tech enabled, interacting with our ecosystem using a digital platform For him, and whether that digital platform should be developed in house, or just white labeled, because it's available off the shelf. Now in either case, in other cases, there are winners and losers and like in any industry, and what I see is that some organization of both natures, so either tech driven or tech enabled, and they are they're changing their business model under way. And they're pivoting and into the other nature. There are several examples on the market. And this strategic move is now where it implies to undertake radical changes. And one has to be really aware of that. I mean, it's a change in skill map, monetization schemes, investments, management, HR, and so on, and so forth. But the good news is that they're great examples of successes in our industry. Quest for data now, artificial intelligence, machine learning, machine learning, deep learning, I mean, all this is now these are our words that are generally generating a lot of money on for their own because everybody's using them. In the end of the day, my view is very pragmatically, I mean, we know that machine translation is now translation, its statistics. And its probabilities relying on large amount of, of datasets. So humans are used to increase the between bracket quality, that means how relevant those datasets are for translation of communication. The point is, those same datasets can be used for other purposes, optimizing them for different market needs, and leveraging on existing delivery processes. Looking at existing market players, how they do in this industry and in other industries, what they're doing with data set, boundary, our data sets are extremely of extremely high quality, because they're very well structured, they are annotated, they are tagged. So this is all added value. And the last to evaluate making the strategic move and creating new monetization schemes. But what I think is that they're relevant examples of successes, you just need to ask Lionbridge his shareholders, what they think about datasets and how they can be leveraged from one vertical turn on. Now for I would say more mundane or pragmatical aspects of how to compete well or better cost cutting. technology reduces the number of labor intensive tasks, I mean, primarily it is deployed to reduce operating costs, I mean, one understand that very well. So to sustain market pressure, we all know there are a lot of different market pressures, it needs to be integrated into a holistic approach to business effectiveness. And that effectiveness must be monitored and must be measured against relevant KPIs. So the focus is to run the business according to Lean management practices. It is quite well described in Reese book, really, and I invite you to go through that as well. Using an analogy with a weight loss program. I mean, obviously, if you want to lose weight, and you know you, you reduce your your weight by 30 pounds in six weeks, obviously, it's going to have some impact on your body and probably negative impact. And that's hardcore cost cutting when when you react as opposed to anticipate. What I do, you can also do is eat healthy every day and practice, you know, workout three times a week, no addiction, and you're gonna be vital and energetic, and be most probably very much more resilient than the other person. So, practically, and I know it's going to generate some some questions, some some even some five rounds. But I think one key thing is to lower project management and engineering costs to the strict minimum or zero it. So we can speak about that at some point if you want or in another session in the future. But obviously, technology side effect is, is the impact on on labor intensive tasks. And that thing when when you're facing changes in your strategy, or you really want to, I would say reinvent yourself is that whilst this process is important is far too often, market players are relying on two or three or four key customers needed sometimes only one. And although it may sound very cliche, new business acquisition is vital and especially when you privatize. It should be a religion really to make sure that you constantly bring in new customers and You want to maintain existing customers, that's for sure that's true. But if you only use that as a tactic to, to grow your top line, I mean, it's an impediment to your strategic move, you don't have any options to test out new products, win with new people with with different people with different persona with different market segments and and you're going circle. And I've seen that, and I've seen that in organizations that are very mature. So moving forward, is the word invest investment. To maintain or create a competitive advantage, you must invest in research and investment budgets, they are not just for larger organizations on the country, smaller LSPs and entity. They used creative ways to prototype their new value proposition and test them out mean it requires financial means and human capital. Unknown Speaker 11:02 Partnerships with academic institutions, national and international subsidies, private equity placements, and a lot of ways to put together investment plans. A good government's advice that I would put on the table is allocate a double digit percentage percentage of your profits on investments, that's also a very wise thing to do. You're creating a virtuous circle, and you're nurturing your own competitive advantage. Last but not least, their translation has shifted from art to science. fluence is replaced by predictable, high end is replaced by commodity or core activities are impacted and there is no way back. We may strive for the best quality, which I respect very much. But it is clear that there has been a shift and this shift will continue to focus. And usability is is the word that that comes to mind here. So it's a great opportunity to think and make changes. And it's a great opportunity for entrepreneurs and intrapreneurs. To leverage on their what I think unbeli is the best quality of those people, it's that they can take risks there. So what I invite you to do is to listed there, and, and private whenever you believe it's time to do things differently. That's it. Thank you. Max Morkovkin 12:48 Well, I think that it's time to move to the questions. It was a short but very deep thought presentation. Thank you very much, Mark. So, let me start with the question from Matias Caesar. NMT is no statistics and probabilities. Did you not find a great shift in how NMT is now available, integrated and used in TMS especially in Agile software development? We're just getting started, I believe. Marc Jonckers 13:29 First, I It's nice to see you Mateus. It's been a long time. To me, it's still by I wouldn't agree conceptually. I mean, the notion I did not say NMT I said Mt. And I said MT as a concept. Obviously, it's based on an algorithm and to me the process of trans translation mental translation process is not the same as the one we use by human person and human and you know, using its cognitive capabilities and and the the best possible output or throughput of an algorithm. So the different that it's that it's just the beginning. Yeah, probably. But that's also in line with what I said that just said that there is no way back and now technology and different ways of different business model this different value proposition This is the future for LSPs in in, in any event, that's my take. So that could be a topic to be discussed further on. But I'll say on my on my vision their. Max Morkovkin 14:37 Mark, can they ask you to stop sharing the screen so that our attendees just see on the screen? Do I Okay, yeah, just up sharing Unknown Speaker 14:47 that. Okay. Max Morkovkin 14:49 Yeah. All right. It's not staggered much. They're moving to the next question. And just to remind the attendees, Mark will be given away Look to someone the question he will like, right? I don't know, what are the parameters for which isn't the best question mark, maybe you can tell. Unknown Speaker 15:09 I go to them because it's out of respect, I very much appreciate those people who are asking questions who are putting the remarks and in front of my eyes, I mean, very, I'm very thankful for that. What strategies can LSPs employ in order to thrive among low priced competition? What the opposite is, this is, you know, we all know that. Well, we all know, Porter as identified, you know, leadership by by cost or differentiation. In any event, you could decide that you want to go for the bloodbath and and really play that low price competition and be the cheapest, you may decide that you want to go and then provide premium services. And it's a choice strategic choice. I mean, if we this book, this blue ocean strategy, although it's a bit complex, I think we could be simplified a little bit my humble take. There are many ways but I would, I would really recommend to not to go for it for me to strategies and doing what exactly other people are doing. That's the best way just to be, you know, drowning in the midst of competition. So that's, that's my take. I'm not sure it's the best answer, but that's the one I have in mind for now. Max Morkovkin 16:27 Try to be unique. Okay, Marc Jonckers 16:29 so shall you? How can I now speak using all these streamlining cost cutting measures ensure that they're trying to human beings who actually communicate and feel feel valued? Yeah, I think value will be will be produced differently than in the past. That's what I think in the first place, that we shall respect to what these people are doing. Yes, for sure. But how they how they will still be relevant in the future will be dependent on the value that they can bring along to the to the automated process that that exists upstream. It's a change I myself am a translator, I mean, I'm a linguist, I I've seen over time how, how this, this, this, this core activity has changed and just invite people to to align and adapt and create new value. Do business value? That's that's something very important. I mean, that's one that's one. So which reason? Yeah, I've tried to answer to all of them, but they're piling up. So I'm glad that you're I won't be able to to answer all of them. And I don't want to take up too much too much. They Max Morkovkin 17:41 have to mark we have time, we have 10 minutes. Marc Jonckers 17:46 Okay, let me processes and then the way i can i can so what region do you think are the best to offshoring right now? Well, I'm a bit harsh because I say no shoring at all, I just say, I think you should keep that's my humble take, you should keep a core team very close to you. It's, we now we are in a period of social distancing. But it's very, very important, especially for a startup or scale up that's going through change that, that this change, we also evaluated not only from from a KPI perspective, but also from a human human relationship perspective. So I would go for proximity in the first place. You can you can also look for for outsourcing. I mean, you have to identify those core activities, you know, you want to keep knowledge of that. And the rest is a choice. But yeah, so Matthias says, the best quality service delivered that satisfy the customer needs 100% that just Max Morkovkin 18:43 the second tried to get the book. Marc Jonckers 18:47 out, he's very smart guy. I mean, we've been doing business together for I don't know about a decade. So. So he's challenging me now. But he's a very smart person. So to me, Wow, linguistic quality, I've heard many things and maybe I'm gonna shock many people. But for me, it's reusability 99% of the content that is translated on a market is as a lifecycle of probably a week or two. So we have to be careful about that. And segment the content that requires specific linguistic, I would say cosmetic approach and and the other one that's that's just it's good for what it takes and that's it. Because there is no there is no economic logic behind that it doesn't work. You cannot spend money and no, the organization or institution will spend money over time on on a content that has no future for itself within bracket. With sorry, with regards to I'm losing a little bit track, but I'm getting there. I mean, you mentioned a cardio you say you mentioned reducing eliminating pm and engineering makes perfect sense aside from automating content normalization And do you see what changing this? Well, project management? Well, I mean, we know it's about file management, let's, let's be honest, let's be, let's be very, very candid. I mean, engineering. I mean, I've had the opportunity to work with mechanical engineers in a, in a, in a startup project that is doing, you know, something that's that had nothing to do with service delivery. These guys, they, they're conceiving the designing very sophisticated processes or machinery. We are looking at five format, we are looking at translatable strings, we are looking at an embedded graphics UI. And that's it. I mean, I would recommend first to disregard a little bit this this, the terminology was we're using and looking at what people are doing. To me, technology can take over a lot, a lot of what is being done in the past and project management and engineering. Max Morkovkin 21:01 Mark, I'm just marking the questions that we answered. So we missed two of them from Iran, if you see how do you balance cost cutting with misleading quality? Marc Jonckers 21:15 Me? Oh, yeah, I'm just implemented. Yeah. Well, that is a good question. It's simple. Few words, not like me. And I think my humble take on this is, it depends on what you want to achieve. You have to have, you know, you always working with the end in mind, what did you want to build what you want to achieve quality you have to define? And you have to make it an objective, will objectify viable criteria? Not just it's good, or alike? Or? It's correct. You need the measurements for that. And then, of course, you may need to decide also what kind of gross margin you want to make on it. And then you can identify how much gas you can allocate. And when you have how much gas you can allocate, and you define whether it's possible to do it using automation, or a mix between automation and, and human in the loop. Max Morkovkin 22:19 Good, then there is a question from market community from York. Bear sec. How will machine translation affect the professional translators or announcements? Marc Jonckers 22:29 Okay, I'm going to be extremely candid here. very candid. And it's not something that I say, because I like it, I hate this idea. But today, we were talking about natural language processing right. Now, some organization have started working on non template based natural language generation. Few guys, most of you are aware that if you can automatically generate a text, you don't need translation in a model you. So the point is, at some, at some point in time hardware, because it's hardware that is making it difficult not software, will enable to process a certain amount of content or concept of designs, descriptions that will be automatically generated. And so that part of the way we take the data, or the content we translated will just vanish into the new. So that's a part of it. And the other. It really depends. I mean, I obviously don't know what's gonna happen in the future for everyone. But Max Morkovkin 23:46 yeah, so the question from Anna Marie roots, reducing pm effort to zero, that seems impossible. I would love to hear more details about Marc Jonckers 23:55 this. Okay. That's a good one on my, to me that there is a prerequisite for that. And again, that's based on on how LSPs position themselves on the market. Do you want to customize all your processes? Or do you want to standardize your process? Do you want to become an organization or a person that is selling between practice standard services or customer service visit? If it's standard? I tell you, it is possible. totally possible. It's just a matter of standardization. That's it. That that we don't want to standardize our processes because we want to offer a different or specific customer experience. Yeah, fair enough. Fair enough. Then know Max Morkovkin 24:42 the question from Habib, donate, what advice would you give to newly translators or so on regarding your domain? Marc Jonckers 24:58 I would, I would I would tell them, expand your skill set, go and become a subject matter expert. And that will help a lot in into providing you more work. Because your expertise, know how that has, that's something different than just your economics is clear? Max Morkovkin 25:24 Hopefully, and there is a question from the agency, I guess AB try to Cherry have you start using or implementing GP three or GPT, three or Google trained 6 trillion parameter ala which model? Marc Jonckers 25:39 Well, we are talking here about data set and and and and algorithms that are available. You can implement them, you can use them, I think what is important? And I would say the second boundaries of the ball in your question is, do you want to build your own models? You want to build your own empty environment? Or NMT? environment? I would say definitely not. I would not recommend anybody to build its own and empty and empty, empty, because they're available on the market. And as you pointed out here, if you want to take advantage of existing content or data set, well, they're there for free, you can use it that are they clean? How good are they in terms of usability? And and are they anonymized? And are they applicable? Again, are they applicable for the verticals you're targeting? To me, you can you can have a Ferrari, I have no problem. I love Ferraris of Porsche, I love whatever. But if I go to the grocery store, I need a car that can you know, bring my grocery back home, without you know, spilling all the milk come on the back of the seat. So it's just again, what you used to do what? Max Morkovkin 27:07 Okay, and the question from not yet or not yet, do you use empty in the Arabic market as much as you do in other markets? If yes, is it as effective as in other languages? Marc Jonckers 27:24 I don't, okay, it's a good point. And I get the I don't have data to to provide you evidence about that. What I can tell you that is that personally, and again, I'm I'm not an expert. I know my deal. But I'm not an expert. But I don't see any reasons why why MT would be less efficient. In one language or another, it's a matter of datasets, it's the matter of how much data is available quality data means, you know, relevant and well structured data. But over time, it's a matter of time. It's just a matter of time. We are generating generating content every day. So that's, that's it. Max Morkovkin 28:11 Mark, thank you very much for the great presentation, we still have questions and attendees challenging you. So you will have to choose one a bit later, I guess, to look at all of these questions, and then choose the one who will share the book. And yeah, thank you very much. It was a great one. Marc Jonckers 28:31 Thank you. Thank you. My pleasure. It was a privilege to be there. And thank you all for the great questions. I'll pick up those who want to book of course, because I know some some either have read it or just simply don't want to get it as I say Max Morkovkin 28:47 I sent you the LinkedIn invite. So Marc Jonckers 28:48 I hope all right, thank you very much, Max. Take good care. You too. Max Morkovkin 28:52 Enjoy your day, too.

Discover why 25% of the Fortune 500 choose Smartcat