Episode Transcript
[00:00:03] Speaker A: Hello and welcome to Localization Today, the stage where we explore how language, technology and community converge to unlock ideas for everyone everywhere. I'm Eddie Arrieta, CEO here at Multilingual Media. And today we're diving into Forrester's first ever wave for translation management systems and what it means for the future of language technology. We are joined by Georg Ell, second CEO of Frase, and Jason Hemingway, the company's cmo. Frase was recognized as a leader in this inaugural wave, earning top scores in 21 out of 26 categories. We'll unpack why this matters, where the market is heading, and how enterprises can turn multilingual content into measurable growth. Georg, Jason, welcome and thanks for being here again.
[00:00:56] Speaker B: Thank you.
[00:00:58] Speaker A: And all right. Of course, I'm not going to introduce you in further detail because you've been here in the show before. Our audience knows both of you. We will put some of those descriptions within the descriptions of this episode, which you're probably listening to on Spotify, Apple podcasts, or maybe watching on YouTube or maybe listening to a snippet on LinkedIn. Of course, this is the first time Forrester has published a wave dedicated to language technology.
Jason, starting with you, why is this milestone.
Why is this milestone important right now?
[00:01:31] Speaker C: Yeah, thanks, Eddie. I think it is a milestone and it's hugely important for our industry because if you look at Forrester, they're one of the most influential analyst firms out there analyzing technology across an enterprise. So actually, to get this on their map is quite a coup because it actually means that, you know, it's a strategic technology now. So Forrester recognizing it means that this becomes not a tool. It's now a strategic technology that all enterprises should be considering as part of their enterprise stack.
And the waves themselves matter because they're kind of this independent, very rigorous process, as Georg and I both know from going through it. You know, a comparison of all the top players in a market, a given market or category, and it's not just grounded on kind of features, although that's a big part. So they review your current offering. So the product that you have, they also look at your strategy as well. So what does your vision look like? What does your roadmap look like? How do you think about innovation? And then the last bit of it really is they look at customer feedback. That's a really important thing. So what are your customers saying about you when you're not sort of in the room? So they interview some of your customers. So it's a reference point for anyone looking in Technology. So number one, it's a reference point that puts technology on the map because for us to put their lens on it, then also it's, you know, if anyone is looking at that technology, they've got a map of all the leaders and kind of how they shape up against strategy and current offering. So being recognized as a leader is a real good accolade for us and generally for the market. The wave is a validation of the technology that's being put out there as being kind of an enterprise level strategic component of any business.
[00:03:10] Speaker A: Thank you, Jason. And what does this tell us about the rising importance of multilingual content for global enterprises across product, customer experience and growth?
[00:03:20] Speaker B: Hey, I'll jump in. I think it's a great question. And the honest answer is that, you know, the combination of AI with the data assets that have been built and the increasingly sophisticated workflows that you can put around that to make sure you're actually getting good value out, the AI and quality assessment and so on open up massive new areas of opportunity for companies with multilingual content. And not just multilingual, of course, now multimodal content as well, video and audio and so on. And that could take in the form of real time customer experiences of reviews that are done in multiple languages, also fine tuning and adapting the content to different audiences, even in the same language. So, for example, saw a customer who did hotel reviews aimed at my generation of British English speakers and then aimed at a younger generation of Australian English speakers. And unfortunately I was totally unable to understand a single word of the Australian review, even though the source content was exactly the same and it's the same language. It just didn't make sense to me because I'm getting old.
But that's fine. That's the sort of adaptation that brands want and need because they have different Personas, they have different audiences, and they want to be able to speak to them very precisely. So that's a huge opportunity that companies have just starting to scratch the edge of.
[00:04:33] Speaker A: And I love that, Jason, because it points to the essence of the recognition rather than the superficial elements of it. And Georg, of course, Jason pointed to the fact that this is a great accolade and that, you know, there are some vanity elements out of this, but when we get to the essence of it, Georg, how does this placement really influence your conversations with your customers, with your prospects, with your partners?
[00:04:59] Speaker B: Well, I think, as Jason has said, Forrester is a very credible analyst firm that for decades has evaluated categories of enterprise software. So it's fantastic to see Forrester recognizing language technology as a critical component of enterprise software for companies.
And so I think it actually gives a lot of ammunition to localization teams to explain the value of what they're doing internally. And I think it helps the industry to explain itself as well. So, overall, I'm extremely optimistic about the future of language technology.
You know, our placement within this report is obviously fantastic. We're very pleased with it. But I think the simple existence of this report as a piece of collateral that people in enterprises can show internally is actually really valuable.
[00:05:47] Speaker A: And of course, Georg, it's 26 out of 21 out of 26 different elements in there. And those categories are really important, of course, but which ones you. You think leaders should care about the most?
[00:06:04] Speaker B: It's just to dive into the scoring a little bit. 26 categories and you get graded below with your peer group or above the peer group. And so actually, in 21 out of 26, we were graded above the peer group, and in the other five, we were graded as with the peer group.
And obviously we're very pleased with that. I think it highlights the breadth of the phrase platform and the fact that we are more than just a piece of software and people use many different components of our platform. You know, we are more than just a component. Right. We are the sum of all the parts. It is a platform. And I think that comes through in that breadth of offering, which I'm really pleased with. So if I was going to call one thing out, it would be on the theme of innovation, which I think we are specifically called out for in the. In the text of the report and in the scoring. We've obviously put a lot of work into innovation. I think we've received some positive reviews on that from various analyst firms over the years and continue to believe that's critically important because technology is changing very fast. So we want to be at the forefront of those changes and bringing those to our customers.
[00:07:08] Speaker A: And Georg, added to the great scores and positioning that Frase has there.
Forrester also praised your vision and the transparency of your roadmap. How are you internally, or as you talk to your customer, preparing them for an explosion in multilingual content in terms of capacity, governance, change of management and all of that.
[00:07:32] Speaker B: Yeah, I think the first thing is that we share the vision very publicly, so we're very transparent about how we view the future of this. We're very transparent about our roadmap and strategy.
We're very transparent about how we want to work with partners to bring it to life and enterprises and Trying to build a lot of trust across the industry. So I think that's the first thing. The second thing is exactly this point, we're working in collaboration with the whole ecosystem. We have the pride in what we do, but also the humility to understand that phrase, as broad as it is, is not the alpha and omega. It's not the only piece of technology, it's not the only people are still going to need some services. So we want all of these things to integrate with phrase and for us to be a core constituent infrastructure component of a customer's enterprise software stack around content and multilingual content automation. So I think talking about it a lot, I think the fact that we have 90 day delivery cycles on innovations so customers can have faith in, we kind of do what we say we're going to because we iterate quickly rather than trying to make big multi year bets about things.
We're able to pivot quickly.
And then of course lots of emphasis on orchestration as workflow so customers can build very custom solutions that match their business requirements. So the custom workflow, the orchestration of all of the different components I think allows customers to experiment and then to put into production those experiments that are successful.
[00:08:59] Speaker A: And clearly artificial intelligence is at the center of such innovations of that roadmap and that evolution of the technologies that are in there. How do you see AI powered language technology evolving over the next three years? Personalization, You've mentioned agentic workflows which from what I'm hearing in our conversations is not delivering as much as people were hoping to do. So there's a lot to a lot of work to do there. How do you see that evolution?
[00:09:28] Speaker B: Well, I think it's really important to. I think it's actually very helpful that the world is currently going through a little bit of a correction on the AI hype cycle. So you know, AI was clearly enormously overhyped.
MIT has put out this study which is now being very widely quoted that 95% of Gen AI experiments in enterprises fail. Interestingly, you're twice as likely to succeed if you buy in the technology rather than try and build it yourself. And I think one of the reasons for that is that MIT say the most successful technologies, AI experiments, sorry are when you are working on a specific problem rather than a very general problem.
So this is what we do, right? We bring the specific application of AI to a very particular use case and then we build scaffolding and infrastructure around that with deterministic software that's looking at quality control data and assets the workflow and put all of that together so that we improve the likelihood of success of these gen AI experiments. So I think we have to recognize the world is going through a correction that is a healthy correction. And people are starting to recognize that enterprise software is not dead, but deterministic software and probabilistic software need to work together.
And I see that as the future of this. I think AI is best when it deals with very, very large data sets and when you have very, very large data.
You know, the physics of chaos theory says that even small variations in the decimal points that you can't take account of unfortunately lead to wild variations in the output. And that is an important thing to say. There's a lot of business executives out there that have a perfectly good translation experience on their phone. One to one, them and their phone and it's great, but that's my, they're my kids, by the way.
But you know, that does not scale to billions of interactions that an enterprise has. And that's where this point around the decimal points you can't capture in chaos theory and the scale at which AI operates in the enterprise needs the scaffolding around it because the experience that you have one to one on your phone does not scale. And that's, you know, that's a very important thing for people to understand as they, as we go forward.
[00:11:34] Speaker A: Love really makes me think about the stock market and the whole correction. I'm pretty sure that's a concept that we want to look into at multilingual. Just a hint there to our writers. But of course, of course there is a correction and it determines a lot of the things that you're doing internally. Was great to hear that you have a 90 day delivery estimate, Jason. You know, the roadmap looks great. You've been praised for that roadmap. It emphasizes hyper personalization and automation. What are the near term innovations that you are most excited about at this point?
[00:12:10] Speaker C: Well, look, those two concepts are the thing that kind of guides us quite heavily and that's just probably good idea to talk about them in a little bit more detail because that's sort of what I'm excited about as a marketer. Obviously personalization has been around for ages, but the idea that you can combine hyper personalization with automation is a really, really important thing. It means using AI not to just kind of translate context content from, you know, a direct translation, but you can kind of adapt it to give customers almost in real time. That's kind of our vision of a personalized experience wherever they Interact. So in the content that they have on the websites they visit, it could be in their customer service interactions they have, but really you want them to be speaking in the language that they're used to, their first language, whenever they interact. And that's how you bring in things like context and that's where things like LLMs can really, really help. You can get into that nuance of what's the context of that person, what's the tone that you're talking to them, what type of audience are they, are they Gen Z, are they older, you know, that kind of thing where you can really tonally adjust things and make it kind of hyper personalized. So we'll be doing much more on that in the roadmap. And that's kind of one of the guiding things. But the other side of it, of course, and it goes hand in hand is how do you orchestrate the right kind of activity around that? So how do you have the checks and balances so that as Georg said, you're sending the wrong thing to the wrong people or putting translations that you've done, not to a human, when actually they do need a human review at the end? Or how do you get the AI to review and send it to the right place? So workflow and orchestration is massively important and you'll see much more come from us on that. The other side of it, and we alluded to it a little bit before, is that kind of at the agents and I think they have a big part to play because they start to bundle things together to make it easy for people to use. Now there are inherent risks in doing that kind of thing. When you're bundling agents together and you're not quite, they're not as reliable, you kind of compound your errors. And that's where, as Giorg quite rightly said, that's where you need guide rails, you need people that have done it, that people that have really experimented with it, to actually use things like AgentIQ in a way that is going to be meaningful and deliver value. Albeit, yeah, happy to experiment. But when you go into a production mode, if you're an enterprise, you cannot afford to be experimenting all the time. You need to make sure that there's acceptable levels of risk when you're doing these things. So AgentIQ of course has a big part to play, but they need to work. You need to have the right LLMs, you need to have the right checks and balances, the right workflow. So these are the sort of things that we're working on to strive towards that kind of hyper personalization, much more automation and making that easy to use and easy to follow, to make things transparent and go in the right places. So, and that all needs things that we discuss all the time when we talk to people, things like quality checks, you know, what's the context of the person interacting? We're bringing out this idea of content profiles which will come up as well soon, which is, you know, what is that audience that you're trying to get to? And it moves away from just checking things, you know, are they word for word perfect as opposed to what are you trying to do with that content, who are you trying to engage with? And that's where more metrics will come in around engaging with content, you know, rather than just saying is it a direct translation? It'll be well what, what is the sentiment of what you're trying to do? So much more on those kind of things that will come up in the future releases.
[00:15:24] Speaker A: All exciting stuff, very exciting stuff. Jason and I hope, of course you remember multilingual to let our entire audience know what's going on there. And of course it's very rapid innovation.
And of course you want to make sure that as many of your customers as possible are adopting these innovations in, in a risk free environment, let's put it that way, or in a way that maximizes the benefits rather than amplifying the risks. How do you balance that rapid innovation with helping them, your customers, especially the smaller teams to absorb to adopt these new capabilities effectively?
[00:16:02] Speaker B: Yeah, I think, I mean just very practically we have, particularly for smaller customers we have an academy and customers can get access to a ton of online training through that.
We also have a customer success team that scales with the size of customer as to the amount of interaction. And of course for the very biggest customers, we'll be heavily involved in helping them. So it's a scalable model, it works well. And then the thing we're trying to build out on top of that is an ecosystem of partners. So we've done things like co investing in solution Architects with LSPs in Europe, in America and in Asia. All of which are ways to try and extend the different ways in which customers can actually get access to really strong phrase expertise.
[00:16:41] Speaker C: Yeah. And I think just just to add to it, you know, the pace of innovation has to be fast because that's what customers are expecting. They're the questions they're asking of their providers is, you know, new things are coming up. How are you reacting to that? You have to have concrete answers to those questions and have a, you know, a robust roadmap that whilst whilst being flexible, has to really put your stake in the ground of what are we trying to deliver, where are we going? And if you can communicate that customers, they absolutely know it and actually they demand a lot of that when you get into the enterprise. So there's that. And I think the one thing I'd also add to what Joerg said is kind of what you're always trying to do is build a platform that is accessible, customizable to each individual business needs. So there's a lot of focus when we develop things on what is the ultimate use case for that, what value is it going to deliver to the customer? And that's something you always have to keep your eye on because there are so many bells and whistles that you could develop. But what's the point if it's not delivering the customer value they expect?
[00:17:37] Speaker A: And thank you Jason, because you point to value and you point to customer and partner value. Georg, if you can tell us a little bit about automation outcomes. Customers have cited automation of up to 99% of workflows. Can you share some examples of how that transform outcomes, cycle times, release, cadence, cost to serve quality, et cetera?
[00:18:01] Speaker B: Yeah, actually coincidentally today we released a fantastic case study with a cryptocurrency platform called Kucoin. So that's K U C O I N. So if you look up phrase.com, kucoin, you'll, you'll find the case study which gives it's, it's actually a really nice case study. It's very well written because it gives a data on the benefits in time savings, in dollar savings and it also speaks about the problems that this customer was trying to resolve. And, and it does talk a little bit about how, you know, what is the actual solution. So it's actually as a just a practical, pragmatic guide for people in localization to have some inspiration for what can be done and how to do it. It's a really good document so I'd recommend people look at that. And the savings can be dramatic. The customer talks about half a million dollar savings per year just in one particular area. We have another customer where we combined quality performance scoring. So this is the quality assessment model that we've built with Fraser Orchestrator for workflows and by setting different thresholds for different language pairs and content types and tweaking those to think about for that particular scenario, we will publish it straight away if it's over a certain threshold and in other Cases we'll go for human review. They were able to save $100,000 a month in the first language pair they trialed just by, you know, instead of sending everything for review, sending what needed to go for review. And I think, you know, broadly speaking, that is the theme of automation. It is, as the name suggests, automation. It is about removing humans from being so much in the loop and ensuring that they remain in the loop where necessary, but trying to reduce how much is necessary. Therefore things can flow faster and that means, you know, things get reduced in minutes or hours instead of days and weeks. Examples also from Phrase Studio, where previously customers were sending videos out to agencies and getting turnaround in days. In Frase Studio it's minutes and so they're able to get a very fast roa. In one case where a customer is actually selling video advertising on their platform, they said not only was there a cost saving, but there was actually business won as a result of that because the turnaround time previously was too slow and the customers were therefore not placing those adverts and now they are because it's much faster.
So yeah, maybe another good example we just saw, it's top of mind. I saw an article today that talked about the gaming streaming service service Stream.
Stream is now categorizing reviews of games by the language in which the reviews are written. Because what they see is that for certain games, the reviews of the games vary tremendously based on the language of the reviewers. So for example, perhaps English speakers love the game but pick another language speakers aren't enjoying it as much. And that's fantastic information for the gaming industry to be thinking about, which is actually localization can go directly to the public facing reviews that we're going to get on Steam. So lots of different ways that localization done right and automation done right make a massive difference.
[00:21:05] Speaker A: Yeah, this is where we see a lot of opportunity and this relates back to that point of the exponential growth of content across many different sectors and industries. So very exciting times for our industry. And I was really happy to hear about the Academy. I'm probably going to get back to the phrase team about the academy and also all the ecosystem elements that I'm here hearing in the conversation. And I know every time we talk there, there are some of these gems that pop up. So we'll be happy to look more into those and talking a little bit more about that ecosystem. Georg, you also earn top marks for your partner ecosystem. What role do partners play in helping customers succeed at scale at Phrase?
[00:21:47] Speaker B: Well, it's critical. I Think, you know, it's interesting, two years ago, I think I remember speaking to one of our major customers, a big software business and the head of localization there. And I finished the call and at the end I said, I will call the CEO of your LSP so that our teams can come together and we together can work out how we can better serve you. And she was surprised. She said, really? You're going to do that? And I said, of course. So we did. And the surprising thing to me was that she was surprised because it turns out this is an uncommon behavior.
So we have really emphasized over the last few years this concept of we have a role that we're proud of.
Software we're proud of, but the humility to understand that we are not the complete and end to end story for customers, that customers may want to build some of their own technology and then plug that into a workflow through the phrase platform. Customers may have third parties they work with who can build software that adds value. And we want to embrace that and allow them to do it. We're running some experiments at the moment with some LSPs to see, you know, they have technology. How can they plug that into the phrase platform that would allow them to solve specific problems for customers that perhaps the generic solution of the phrase platform doesn't do on its own. And I love that. So what I would really like anyone listening to this who's in the enterprise or in the sort of LSP or the just the tech community, is to understand that Fraser goal is to help everyone be successful, ultimately the customer, the enterprise in the middle.
And if the best solution for that customer is some of our technology and some of their own technology and some of someone else's technology, we see our role as facilitating all of that, working together to the benefit of the customer and not trying to constantly say that everything has to go through us. So, you know, that's how we see the ecosystem. It's a core driver of making customers successful.
[00:23:40] Speaker A: And really well put and really good to know. I think the second time we talk about that specific element on making everyone successful. And I'm really happy to hear that every time from you, Gert.
Of course, Jason, talking about infrastructure, Forrester also mentioned the vendors are beginning to leverage your infrastructure, probably because of this approach that Georg is mentioning. Do you see this as validation, competition? Both.
[00:24:05] Speaker C: Yeah, it's probably a bit of both, to be honest, Eddie, because it is validation, but it's also competition. It's validation because when you point to others starting to leverage the infrastructure that we've built. It confirms that we're kind of our vision and what we're doing is shaping the market. At the same time, it also helps us sharpen that competitive edge, which is good. It's good for us because it means that we have to keep innovating to stay ahead.
And really, ultimately, I guess the good news for customers is that we won't stand still. We've got a leading platform that helps customers scale and we'll continue to push those boundaries, fold in new technologies to the end goal of hyper personalization and that ultimate goal of automating more more and more and more where you can. So I think it's validation and competition. But both are good, right? The competition isn't bad. It's good because it drives us forward, it keeps everybody going forward and ultimately raises the bar for everybody. Right?
[00:25:00] Speaker A: It does. And I think I invite those that are listening to this conversation, if you have not checked the information that's out there around this first wave, that you take a look and that you make your own inferences, that you look at your own insights, then you take your own decisions of what you think it means, please share it with us so that we can also enrich this conversation. Of course, Jason, this says a lot about the market like you're mentioning and where the market is heading. What does this first wave tell us about where language technology is going from the tools to the operating system for global content?
[00:25:37] Speaker C: Yeah, look, I think that's a really interesting point you make an operating system because one of the things that we're very proud of, kind of positioning phrase as is this kind of multilingual content automation platform. It's no longer this idea of niche tools. It's now about strategic enterprise technology and strategic enterprise platform in our case, you know, so I'll go gilg, you've probably got.
[00:26:03] Speaker B: I thought there was a pause. Sorry, Jason. But for me, what it really says is because there's clear blue water between Frasers, the leader in this Forrester wave, and others. And I think I put that down to our investment in the breadth of the platform. Right. So we talked recently, Eddie, about moving from multilingual to multimodal and then from multimodal to multidimensional.
This concept of time and ephemeral content and real time content coming through as well. I think that's what has distinguished phrase in this case. And I think that is a direction finder that is, or an arrow that points in the direction that this industry will go. So I would expect others to follow and that's good because it's the right direction for the industry.
[00:26:44] Speaker A: And of course, we are coming to an end of our conversation.
We'll dig deeper as we get more insights from the Forrester wave. But today I'd love to get your leadership perspectives, if we can get started with you. Georg, what does being a Forrester leader enable you to do differently? In vision, in investment, global impact?
[00:27:07] Speaker B: I'm not sure that we'll. I mean, apart from, you know, proudly telling customers and partners and prospects about it, I'm not sure that we'll do anything differently because it's a. It's a validation that what we're doing is right.
So it's probably a case of continue to focus on innovation, continue to focus on orchestration, on integrations on the ecosystem, on, you know, bringing specific AI to specific use cases and kind of like narrowing it so that it increases the success rate of these experiments. So I think we're doing many of the right things and we want to scale that up. We want to do more. So I hope that the wave will introduce customers to us that we haven't spoken to before. It will maybe create a context around phrase that will help people to have a really rich understanding of us as they engage with us. So I don't know that we'll do anything differently, but I hope it opens doors for us and allows us to do more.
[00:27:59] Speaker A: Fantastic.
[00:28:00] Speaker C: And just to add Eddie, a tiny bit on that, I think it helps us help the industry elevate conversations in businesses to somewhere that they haven't been. So rather than, you know, there's the old adage always that localization is an afterthought. Well, it doesn't have to be now because you've got a validation from a massive independent analyst firm that says it's not an afterthought. So in that sense, decision makers outside of localization, you know, think, you know, people like me, CMOs or even CIOs or Chief Technology officers can now sort of have something, a reference point. They go, oh, okay, I understand that this is actually should be part of our business and we should give it more thought and actually give it more of a seat at the strategic table. So I think it helps in that sense and helps frame it as part of a growth agenda for any business. So that's kind of my two pence on that.
[00:28:47] Speaker A: Absolutely, Absolutely, Jason. And I personally hope that these helps our industry elevate the conversation, approach it from different angles. And there is, of course, an opportunity here to communicate with business leaders. If Both of you could let me know from your perspective, if there's one message you want business leaders to take from this recognition, what is that message? And I think we can read between the lines from our conversation here, but if you can explicitly mention it to them and also, what should they do in the next 90 days? I mean, it sounds like that's your next delivery cycle. I don't know where are you in those 90 days? But it's a great opportunity for everyone to take on that 90 day cycle too.
[00:29:29] Speaker B: Yeah, we do four releases a year. The next one is in three weeks, I believe. So we're excited for what's coming up there. And then the next one will be in December and then and so on from there. I think the answer to the question is a little bit of a summary of the whole conversation, which is to say, as Jason just said it, this report really validates the space and it validates the opportunity that exists for enterprises who lean into it. And I gave some examples around gaming. So time to market can be massively increased. You've got the opportunity to launch your product immediately in 20 languages or 50 languages instead of starting with one and then slowly, you know, adding more languages over time. You've got the opportunity to do marketing campaigns in 30 languages every week instead of English first. And six weeks later, you know, 10 languages in six weeks later and the opportunity's gone. So many companies I speak to, even the biggest ones, even like trillion dollar market cap companies, still have big divisions doing localization on spreadsheets and with salespeople. And that is big bonkers technology exists to resolve this and also to offer entirely new customer experiences. So I think hopefully this Forrester document will validate the space to Jason's point, give business leaders the insight to say, right, this should actually be something we look at standalone rather than as something that is a, as I said earlier, you know, solved problem, which it is not. It's complex. But the opportunities and the benefits there, if you grab them.
[00:30:58] Speaker C: Yeah. And my only advice really would be to take a look at it and evaluate where you are as a business. Where are you, where are you against this kind of benchmarking that Forrester have kind of done?
[00:31:08] Speaker A: Fantastic. And of course that signs the end of our conversation. Gjerg, Jason, thank you so much for joining us today.
[00:31:16] Speaker B: Thank you.
[00:31:17] Speaker C: Pleasure.
[00:31:20] Speaker A: All right. And that was our conversation with Georg El, CEO of Phrase, and Jason Hemingway, the CMO of the same organization.
As you have heard, we have talked today about Forrester's first ever wave for translation management systems and what it all means for the future of language technology. We will continue following up this story, your insights and comments on the podcast that we will be publishing on Spotify, also Apple Podcasts and all of the social media that you know. Please comment, share and follow us. My name is Eddie Arieta. I'm the CEO here at Multilingual Media. Until next time, goodbye.
