Finding Automated Localization (And How Jordan Lewis At Culture Amp Talks About Translation Internally)

podcast

It’s standard to keep a finger on the pulse of our industries and competition. But can you say the same for how you pay attention to your employees? Have you examined the pain points of your own company? Do you have a high turnover rate?

Culture Amp is best known for helping companies improve employee retention through facilitating engagement and peak performance by collecting feedback via simple surveys. Culture Amp turns these insights into high-impact actions to create an ideal employee experience.

In this episode, Jordan Lewis, Director of Product (Platform) at Culture Amp, sits down with us to discuss the internal and external growth that continuous localization has enabled for his company. Jordan shares how his team at Culture Amp drastically improved translation turnaround time (from four weeks to a few days) and how his team leverages Slack to not only communicate with translators to reach results faster, but also to keep the entire company up-to-speed on the localization strategy.

Subscribe: Apple Podcast / Google Play / Spotify / Stitcher

On this episode you will learn:

  • How to implement continuous localization into your product development cycles and reduce turnaround time from four weeks to days
  • The best questions to ask yourself as you strategize for translation and localization
  • Ways to tier content and product features to optimize translation spend
  • Why it’s important to have direct communication with translators

What to listen for:

  • [01:55] About Jordan and Culture Amp.
  • [04:47] The impact of COVID-19 on buyer behaviour and engagement.
  • [05:45] How the new translation initiative varies from the original.
  • [07:28] How Jordan’s technical understanding informed the creation of the translation management system at Culture Amp.
  • [11:00] Did the end user experience impact the creation of the system?
  • [12:45] Understanding which markets and languages to target for translation.
  • [14:45] Ease of adoption of the system.
  • [16:45} What process of translation review does Culture Amp have in place?
  • [19:24] What surprises Jordan faced when creating the translation system.
  • [21:55] Evaluating the efficacy of the system.
  • [23:55] What the future developments or implementations at Culture Amp could look like.
  • [25:40] What the current translation process at Culture Amp is.
  • [28:43] What kind of automation is in place for content reaching correct workflows?
  • [30:25] Talking about localization through the lens of business.
  • [34:13] Common questions Jordan gets about the translation process at a company level.
  • [36:33] Creating an open and informed culture of localization within the company.
  • [38:22] Advice for localization teams working on similar processes.
  • [40:45] Where Culture Amp sits on the Localization Maturity Scale.

Find Jordan Lewis and CultureAmp online:

Full Transcript that almost certainly has typos (forgive us!) Announcer: You're listening to The Loc Show, presented by Smartling.

Adrian Cohn: Hey everybody. Welcome back to The Loc Show. I'm your host Adrian Cohn with Smartling. Thank you for listening. My guest today is Jordan Lewis. He's the director of product platform at Culture Amp. Culture Amp makes it easy to collect, understand and act on employee feedback. I loved this conversation because Jordan is humble about the sophisticated localization program that his team has built. They have a fully automated translation program. They've created three distinct tiers for how widely a language would be supported across all of their products. And they have additional tiers for quality by content type. Isn't that cool? And if that's not enough, they also built a Slack channel for the whole company. So people can check in on localization any time. That's one way to get by in. Right? Let's get right to the episode. I hope you enjoy it. Jordan, it's great to have you on the show. How's it going today?

Jordan Lewis: It's going really good. It's great to be on the show. I'm really excited to be sharing all the stuff we've done with Smartling product and how it's really helped us get better at localization and whole translation process. So it's good to be here.

Adrian Cohn: We've been anticipating this conversation because this podcast exists for the purpose of telling stories of people like you. And you are running the product team at Culture Amp, is that right?

Jordan Lewis: I am a director of product for the platform camp. So camps are our departments are our groups. So platform is a 40 odd app campus which is our name for it.

Adrian Cohn: Good.

Jordan Lewis: And we sort of look after the product where we've got, we own the shared product and also the shared experience. So localization cuts across all our products and it's something I've been looking after since I started well about early last year. So I've been at Culture Amp now since January 2019. And it was the first thing that I actually got tasks with as when I joined was to fix up our process. So I've been involved, but I've also been at a bit of an arm's length as well at the same time, once we set it up, it's been running very smoothly.

Adrian Cohn: So nice. Well, we'll certainly dive into that. Tell us a little bit about Culture Amp. What is the platform do and what sort of problem are you trying to solve?

Jordan Lewis: Cool. So Culture Amp, it's a people and culture platform and we're one of the leading providers. So what really do is help companies take action to improve employee engagement, retention, and performance. And we do this through like making it easy to collect, understand, and act on employee feedback. So we help both the organizations, but also the individuals with the organization. We have two product modules that our customers can choose from depending on their needs. So have one product called performance. This helps managers and teams stay connected and focused and make sure that their employees are productive with continuous feedback, goal tracking, and flexible performance reviews and things like that.

Jordan Lewis: On the engagement side, this is more for the organization to have a pulse on how things are going. So this can either be things like surveys throughout the life cycle of an employee like starting and how are they going throughout the year and then all the way to the end where they are off boarding. But it allows us to also do like a pulse surveys and engagement surveys, ad hoc surveys as well. And we take all this data and it allows us to get really deep insights and analytics into how the organizations and individuals are going.

Adrian Cohn: What an awesome tool. I mean, I'm so into that sort of discipline of employee engagement, employee satisfaction, peer learning, learning from one another and understanding not just from an anecdotal level, but from a statistical level, like what do people have to focus on? I think it's a really cool problem that you all are solving. And I suppose now with everybody working remotely or predominantly remotely, it's become more important to have a tool like this in place. Have you all seen a change in buyer behavior or engagement with the platform?

Jordan Lewis: So I think if you're referring to like the COVID and the pandemic we're in at the moment, we've definitely noticed a change in the needs of our customers and even internally we've actually needed different things. Like one thing we've got better at is doing pulse surveys. So like a weekly survey to check in and versus the big, long quarterly surveys. So we've definitely being able to like pivot and help our customers in their time of need, whether its specific surveys to understand how things are going or returning to work and things like that. So our platform is very flexible and we're being able to adapt quickly. The great thing is getting these new surveys translate it quickly. It's been feeds right into it. So it's been very, very good to be able to get to market quickly with translated surveys, to help people in their time of need.

Adrian Cohn: Awesome. So when you joined Culture Amp, you said that your first task was to solve for localization. When you were tasked with this, was the business already translating the platform, or was this a new initiative that you were to spearhead?

Jordan Lewis: We weren't translating. We were using a very old product and it was quite a manual process. So it was like the frequency between updates was months. So like, as we were developing things, we wouldn't make orders for months. The turnaround time was weeks, when we did place the order, it would take weeks to get them back and then get them into the system. And then it could take an engineer more than 10 hours of work to actually get the surveys back into the system. We did the translation, it was like very clunky and it left a lot of room for human error and patches in like what was translated and wasn't translated.

Jordan Lewis: So on day one, my task was to use my sort of skill set as like I'm an engineer by trade. And when I joined Culture Amp, I switched to a technical product management and was to really like get a translation management system in place to be able to get us to the goal of continuous localization. So where it was hands off, like where engineers wouldn't have to consider what they were doing, PM's wouldn't have to be a bottleneck in the process. And it took us a good part of the year to do a lot of vendor review, understand the problems, and finally get all the way to the point where we actually had an automated solution that now it's pretty hands off and automated.

Adrian Cohn: I's so interesting to me. Did you have localization experience at all before you got into Culture Amp?

Jordan Lewis: On the engineering side? So like setting up all the translation falls and like the replacing hard coded texts with the ability to despite the translation. So nothing to do with the like a linguist or the translation management software. So this was a new experience for me, but once I discovered it, it made so much sense and it was quite obvious that there was like a big missing piece of the puzzle, which was that translation management system that was missing.

Adrian Cohn: Well, that's what I wanted to hear about because you came in and it sounds like you had a fundamental understanding for the importance of internationalization for your product to separate the strings that were in your product from hard-coded text, so that you could more easily swap content. So you already had an idea of how this works technically, and then I guess what happened? You started to investigate different platforms on the market and you created a vision for continuous localization, or had you heard from other people in the space that there was this Holy grail? I mean, I'm just so interested in how people discover the idea of continuous localization.

Jordan Lewis: So I think like when I was running through the shortlist of the vendors, I think I had a desire that we had a translation management system that allowed for automated translation. So if a developer opened up a pull request that would trigger off our translation and so on, and then it would automatically come back. The other part was the linguist side, we could have gone directly to one company to have their linguists translate our content and another company for our translation management system. What I was looking for was a system like a company that offered us the flexibility to have both, but also like there was other, like really requirements underneath, like the quality and being able to integrate it into our QA process and also have two way communications with the linguists.

Jordan Lewis: So it was lots of different options along the scale of like cutting out the middle translation management system and going directly to linguists. So going to one company for one thing and another company for linguists and that really helped us like going on that journey, then we realized, well, I realized that there was like a whole industry that I've never been exposed to in like the idea of continuous localization. Like once I discovered it, I was like, that's exactly what we want. So once I realize it existed, what we needed, it really helped narrow the search.

Adrian Cohn: Yeah. Well, I mean, that makes sense because your company's prior process was... I suspect that you were shipping new product on an ongoing basis, but you weren't able to deploy the localized version of that new product, because you only would submit those orders every couple of months.

Jordan Lewis: Yes. Essentially before I arrived.

Adrian Cohn: Before you arrived. Right. So most of the time when we hear teams talk about that, what we come to appreciate is the user experience was not what your company had expected to deliver. What we hear is that the ability to get your experience to market for all of your different languages and markets was insufficient in terms of what you were hoping to achieve. Is that sort of what the sentiment was before you joined? And that was how it came to be so high on your prioritization list?

Jordan Lewis: I think just trying to cast my mind back, I think like the initial problem was like the process was slow in error prone and evolved a lot of engineering effort. So that was the first thing to solve was just to like, optimize the efficiency of the process. But once we'd done that, it unlocked a whole other, then the next big questions is like, right now, we can actually translate like, what is our strategy for translation? Like what markets, what languages should we offer? How much should we offer for specific languages? Like what parts of our products should we, and shouldn't we translate. I've forgotten the question, but I think like the first step was just to get it better. And then, the next thing is done. It's allowed us to really get specific and have a strategy for how we get into certain markets and what languages we supporting, because it's so easy to order translations. We could easily or more translations than we need. So one of my thoughts lately is, and like, what is our strategy for supporting new and existing languages?

Adrian Cohn: Yeah. That's a pretty important thing to think about and strategize behind. How did that conversation initially go? Were you given a list by the business and said, "Hey, we got to support these five languages." Or did you do some sort of data analysis to understand which markets were performing the best and approach it that way?

Jordan Lewis: When I joined, we had 44 languages that we offered support for in different, like it was patchy. So the UI might not be translated, but the survey templates would be translated. And over the years we've just built up like all these languages that we have offered support at some part of our product. And then what my first task was to was make sense of that. Like they able to tell our sales pipeline out account execs, like this is the lay of the land. This is what products we have in what languages then at the first step you needed to do was actually to create tiers of languages.

Jordan Lewis: Tier one is like languages we want full support for no matter what, like everything should just be in this language. It's a strategic language that we want to support right across the board. And tier two was the languages which historically supported, but we didn't want to remove support for, so I bet we didn't want to roll them out further. And then tier three we had an option for customers to BYO translations for certain things. So it was a matter of making sense of what we had and working out which languages we wanted full support for and how we support the old languages.

Adrian Cohn: It makes perfect sense to me. I mean, it's a very strategic step that a business can take tiering your markets and then aligning your translation strategy around that. And you alluded to this fact earlier is that it's so easy now, it sounds like it's so easy for you to translate content that if you want to make changes to the strategy, is it relatively easy to adapt?

Jordan Lewis: Yes. So we've we took the Smartling product. We couldn't use some of the out of the box just for security reasons. So we built this sort of like this middle, middle layer between like a very lightweight thing that I could get hub repo that we could actually give access to Smartling. So by the time we built this thing, it's a matter of just a conflict file. It's like moving the local ID from not supported to supported. And then we merged that pull requests and then all of a sudden it'll go off and start ordering stuff. So on the simplest level, it's a matter of just updating and configure what we want to support on the other level it's then there's all the questions from the linguists that we need to actually, they're directed to the right area.

Jordan Lewis: And there's a whole process around making sure that we roll this out as it like another product release, like from the moment we decide to add it, that's a little conflict change, but there's a process we need to follow as well. So it can be very easy to add a new language now, but it involves some sort of process management. They wanted to add it. It's just out of sight, out of mind, it'll be automatically ordered each time there's a change in the product. And there might be questions from the linguist we need to direct them. That's about it, where it's at now.

Adrian Cohn: So many companies are product led these days. And sometimes companies have localization teams that are sitting on the product team as well. It sounds like that's really what you've done, where you've led the process of integrating an automated translation system. And I assume you're also using human translation for the actual content. But you brought up the point that from time to time, there are translations that have to be reviewed. There are questions that come up in the process, who's responsible for that?

Jordan Lewis: So we don't have a localization product manager, or like there's no specific person whose job is to manage localization. And I've chatted to other large Australian tech companies where they've got whole teams in like different regions just to manage that. And I think like chatting with them and I won't name names, but like a lot of the automation can we put in place has actually negated the need for having these whole teams at what we've got at the moment. And you might be expecting us to have some fancy process, but really it's hooked up to the questions to Slack. So when they come in, they come to Slack and we can direct them to the right team. So in the conflict with hooked it up. So if a specific product has a question raised by linguists, it'll alert in a specific Slack channel, and it's either the people scientists or engineer or product manager, whoever can ask the question, we'll answer it.

Adrian Cohn: Right. And they respond in Slack. That question then is passed through to the translator via Smartling. And you can resolve the query.

Jordan Lewis: I think they might have to jump into the platform to respond, but it's the question is surfaced to the right team by a Slack.

Adrian Cohn: That is a slick system that you've put in Jordan. So kudos to you.

Jordan Lewis: We had one team and I'm definitely going to stop and shout out. As a team we've got called delivery engineering and they've really done a fantastic job to get this whole process in place. Like I was working on the vendor review process and getting it all set up, but all the technical implementation definitely goes out to the team delivery engineering who's done all the hard work and made it super slick. And like I said, it's like automated the need for whole teams to be focused on this.

Adrian Cohn: That's so cool. That's so cool. You said earlier, Jordan, that you came in and you had this idea that you needed, well, you knew that you needed to work on localization, and then you have this idea that automated translation, continuous localization could be part of the process. And you mentioned that you kept learning more and more and more about the world of translation. What are some of the other things that came up that surprised you? And how did you approach it?

Jordan Lewis: I think for me, when the penny dropped was when I realized that, like I was chatting to the Atlassians localization product manager Melanie Highway, she used these terms, translation management system and language service provider, but she used TMS and LSP. And I had to go away and Google what this meant. And I'm like at the penny dropped, it's like, there's actually these industry terms for these systems that have very specific jobs. And then when I realized, all right, there is a system that is for managing translations and separately, there's the language service provider. Like that was a huge moment for me when I realized things made sense, like what I needed to look for and where we needed to optimize for the other things I learned was really around just the whole, like the strategy for languages we support.

Jordan Lewis: And that's something that I'm still on that journey now of working with our counterparts in sales and account execs on like that whole strategy for what we support, how we support it and so on. And so it's there, there's so many pieces to this and then think localization it's an overloaded term anyway. It can mean a lot of things to different people. So be more specific around which parts we're talking has been really fantastic to help help us on our journey.

Adrian Cohn: So it's understanding the different areas of translation and localization that are relevant to you, and then focusing in on those components?

Jordan Lewis: Yes. Breaking down the whole localization world into like the components. Exactly.

Adrian Cohn: Yeah. I think that's really interesting because it is a big and complex field and there are seemingly an endless number of problems to solve or opportunities to action on. And at the end of the day, you have to prioritize the things that will make the most sense for your business. It's quite possible that for you, continuous localization is most important, but for the folks down the street, there's something else that's more important. And it's not to say one is more valuable than the other, but I am curious how have you looked at your localization program now that you've rolled it out and it's working? How do you evaluate if it's efficacy, are you tracking a translation quality in any way? Are you looking at user behavior differently or is the goal to just ship content in different languages, according to the tiers that you have outlined?

Jordan Lewis: I'd say it's the latter it's like, yep. Big thing was to be able to ship the language is we want to ship and identify the parts of our products that need uplifting to be able to support translations. A part from that there could be another big push and a whole team dedicated to improving quality and like doing more analytics on our localization. And that's something that's on my radar that I'm sort of juggling localization as I, as one part of my role. And that's something I definitely want to get to, but it's like because we've just had so much success with that first automated process. Like it's brought me time to be able to turn my attention back to that.

Adrian Cohn: Yeah. Well, I think it's a good result if you've been able to achieve the first goal successfully and hopefully it gave you some breathing room to focus on some other big projects that I am sure you have been tackling with the same level of success. So you're now in this position where you've got the fully automated solution, you're able to place orders on a continuous basis. And when queries come up, you're able to pass them easily to folks on the team and they're able to respond. What is there anything that you're thinking about working next in this translation journey or are you all sort of taking a pause on development around localization and just continuously supporting the product development?

Jordan Lewis: So my next steps, I'm trying to, it's more around process internally. Like now that we've got three products or there's about over half a dozen code basis hooked up to this automation process. So having a product manager or a point person in each product that we can catch up monthly or fortnightly and review the languages we're adding, where things are at and just getting better at seeing things through to the end, like when we add a new language, we need to make sure that it's completely added in this product and the second product and the third product, and just getting better at sort of making sure that we are seeing things through and that they are product releases when we add a language. So we need to sort of treat them like that and get a little bit more rigor around that.

Jordan Lewis: And the other big focus, like I said, a few times is the strategy for what languages we support and how one of the thing that became obvious quite quickly was that with 43 languages that we had in the past and our word count, we think it's up to around 70,000 words for each language, it could easily, like the cost could be hugely. So pass half a million dollars just to translate the continent we have. So we need a strategy to make sure that we're translating the right thing. They're getting for the right reasons. There's a lot of thinking to be done around that too.

Adrian Cohn: That's a great point. What's the current translation process that you're using? Is it human translation, a two pass translation for every single piece of content? Or have you got multiple translation workflows that may include machine translation or a single translator? What have you been trying?

Jordan Lewis: I think for our tier one, we've got a human translator with a review step. Not quite sure it's been over a year since we set up tier two, but I think has maybe just a no review step such as the human translator. There're different types of content we translate. So it's hard to put a blanket on like certain things we want much higher quality. So how I describe our product is we've got three types of content to translate. One is the UI, some of the things in the interface that we could push a change... If we see something broken, we could push a change and it's pretty quickly the other stuff, we've got a template. We call them templates because our customers will copy a survey template or an email template. Once they've copied it. And the translations with it, we can no longer ever go back and update that. It's in their hands. Like they've taken a clone of it as a point in time. So those are the ones we want super, super high quality on, because like I said, we go back and make fixes, once they have been cloned.

Adrian Cohn: It's like an email. As soon as the emails out, you can't get it back.

Jordan Lewis: And these things live in databases. So, but that was the other big challenge is like, we could easily translate things that lived in code basis, but things that lived in databases, we needed an extra step and that was quite tricky. And the third one is human user generated content and teacher is one thing, but then the topic content within those tiers is another thing. So it's a bit of a matrix going on and it gets very complex very quickly. So we're trying to simplify things.

Adrian Cohn: Yeah. I mean, but as you've alluded to, like localization is complex and it sounds like what you've done is you've actually accomplished quite a bit because you've done an analysis on the markets and then you've done an analysis on the content and you've divided the translation quality, or how much you will pay to get the results of translations based on the value of the content. I mean, that is a sophisticated approach to solving this. You mentioned earlier that it can be an extremely expensive effort to manage, especially if it's not managed carefully. So these different, the way you've segmented the content. Have you also, do you recall, you might not because there was a while ago when you set it up, but have you also created automation so that when content comes from different parts of the platform, it automatically is going to the correct translation workflow?

Jordan Lewis: Also, the workflows, we haven't touched since we set up. I think so. I think at the time we put a bit of effort into like, translating similar languages, like Spanish, Spain, and Latin American Spain, and English, UK, English, and US English. So I think there's a lot of things we set up. Like to be honest, we haven't actually gone back and taken another path, like a lot of the Smartling settings, because its sort of like we've set it up with and what sort of it's working well. So we haven't had a need to go back and they will be probably in the future where we look at when we get around to doing the analytics on things, we can work out if we ever need to do machine translation for certain things, which that's a whole, but it's not on our radar yet, but it's not out of the question that would go back and fine tune it and tweak the settings a bit to make it even more efficient.

Adrian Cohn: That sounds cool. A lot of the time, you said you put effort into the setup and then you don't actually have to make changes unless there's a problem. So it's not uncommon for businesses to set it and forget for a while. So that makes sense. You work on a team, you said there are about 40 people on the team you work with. How do you guys talk about translation as a business unit? Is it a common talking point in team meetings or like, what does it look like?

Jordan Lewis: It's a great question. So the team I work on, we're one, well, we use all these internal comments, like our group, we call it a camp. It's a same as like a Spotify squad or it's like a large enough that it's a team of teams model. And then, we've got three products. So our product group is split into the three products in a few smaller teams. But I think what I've asked and what I think like is a good way of doing it is like localization should be like built into every feature we built, same way we do accessibility, same way we do quality testing, like new features should be translatable from day one. So that's one aspect of the engineering side of things.

Jordan Lewis: And on the other side of things where the conversations happen, it's like when we've got customers or potential deals that rely on specific languages that we don't support or don't support the module they need, then we have a whole conversation around. Okay. Do we support it? Is it a tier one or is it a tier two and have these conversations and there's more of a framework in place for those decision making. So that's probably the most, like that's where we have the most conversations. We've got this company-wide Slack channel called localization or product localization. People from all parts of the business will come in and just throw their questions or answer questions or, and there are follow along. So I'd say it's not a product group thing, platform camp thing. I think it's a company wide interest. And then we funnel all the questions through a Slack channel. And that seems to be the most effective way of having those conversations.

Adrian Cohn: Having the ability to communicate with a translator during the process of localizing your web app, your mobile app, your website, it's essential, it's essential because things are going to come up in the translation process that require your attention. If you're the content creator, you're going to get questions or feedback from translators based on the source content. Maybe they don't fully understand the content that they have to translate, or maybe there's just a typo in the source content that's happened before. Right. So having the ability to communicate with a translator is really important. And what you're hearing from Jordan is a fantastic story about how he's leveraging different technologies at snarling offers that will enable you to communicate more easily with the translator so that the options within smiling about how you can communicate with translators. We've got three. The first is you can communicate with the translator in the dashboard, in the platform, on the string level, super easy.

Adrian Cohn: Another way. My favorite is by integrating with Slack and you'll get your questions from translators into dedicated Slack channels that you set up and you can respond in the Slack channel, which is super cool. And the last way is you can also do it by email all on the string level. So three great ways to communicate with translators. It's the best way to control the quality of your content. Make sure you're getting to market as quickly as you possibly can. And hey, translators are real people. So it's good to have another relationship with somebody somewhere around the world. Let's get back to the show with Jordan. What are some of the type of questions that you get in that Slack channel?

Jordan Lewis: So common ones, which we used to get a lot off until I built a good documentation page. It's like, do we support X? Do we plan to support Y? Like, could we support Z? Questions on languages for our customers that are asking about, and occasionally you'll get questions around, like more technical questions or like status update questions on this language we added last month, how's that going? Or things like that. So they vary. It's usually more on which locales do we plan to support, and could we add, stuff like that.

Adrian Cohn: I've actually never heard of a company that had a Slack channel dedicated to localization that everybody was in and asking questions. Like the ones that you've just shared. That excites me. It excites me a lot because translation and localization is oftentimes like this very nebulous process that a few people, sometimes it's localization teams that a few people manage. And what I've observed is that they have a lot of difficulty engaging the business in their work. And because it's complicated, and because for many organizations, there are translation processes, what yours was like before you started working with Smartling, people really didn't want to get too involved because it got a little crazy and everything was taking a long time and there was no visibility into the process.

Adrian Cohn: You've changed the way that a business can think about localization at Culture Amp, by making it visible to everybody. You solved the problem by realizing continuous localization, but it sounds like you've also created an opportunity for your whole business to become very engaged with what you're working on. And I guess just to bring it back to your product, to Culture Amp, is this methodology that you're using at all, something that you learned from the utilization of Culture Amp as an employee, or has the philosophy of your business helped to facilitate the process that you developed?

Jordan Lewis: I can give you a more recent example of why we recently did our Q1 engagement survey. And one of the big things that came up was working across company boundaries. And then this is exactly the sort of thing that we need to do is to be able to have a process where we can have like engineers talking to sales reps and account executives and product managers jumping in. So this is something that we were consciously did to be able to have these conversations out loud. And we're very transparent company internally, and like having these conversations, I'd never really thought of it until you sort of like might have a lot of these conversations and know how it works in other companies. But for me, it's been quite normal.

Jordan Lewis: But we do look for opportunities to help bridge the gap between different parts of the business and the localization one is there's work to do on like the strategy and the criteria for adding local languages. But yeah, that's been quite normal for me since I've been there. I've never really thought about it. Yep. Yeah. Now like you said, we've got continuous localization. The next challenge is just is working out how we best sort of use that power we've got now and getting everyone's input and then opinion.

Adrian Cohn: Jordan. If you were working with a localization team that didn't have the developers and engineers that you have on your team what sort of advice would you offer to someone who is in that position? How do you recommend they go and work with engineers and product managers to sort of demonstrate the value of integrating an automation technology?

Jordan Lewis: That's a great question. I think from our experience, so say last when I joined from most of the year was focused on like vendor review and getting through legal and security and probably at three, one to two quarters was actually implementing it. And that team has now shifted onto different things. So we actually don't have a localization team, localization PM. It's all being set up and it's running very smoothly. So I'd say point to that and say, look, you don't need a localization team. Like you can have a big push to get it set up, and then it's low touch afterwards. I think the big things that will bite you is like, if you need to make your product support right to left, or you need to actually go through your product and make it support translation. So it depends on what stage of the journey they're at.

Jordan Lewis: I saw a great diagram. I can't remember it, but it's like the stages of localization maturity. And like, I'd say like yeah, identify where you are on that scale. And it'll give you a point is on how to get to the next one. So I think if you Google localization maturity scale, it will come up in Google images. And I think that was a big sort of guiding light of identifying where we are and where we can get to, but where we might want to get to and stop. So I think it all depends on your core basis and your immaturity level. I think my answer would be different.

Adrian Cohn: I love that you brought in the localization maturity model and I'm going to ask now, so the model has five stages, manual, automated, agile, centralized, and expert. Where would you pin Culture Amp on the maturity model?

Jordan Lewis: Sorry, just pulling it up. I haven't looked at it since last year. Where are we? I'm just trying to look through Google. I'd say somewhere between a three and four. So between managed and optimized, like I think lot of the conversation has hinted that we've sort of gotten to a point where we're happy with and we've sort of put our attention elsewhere. So we're optimizing our processes internally, but we could also optimize our analytics. We could also optimize our settings for it and quality. So I think there's probably, between three and four, I'd say, I don't think we ever want to get to a point of diminishing returns going any further than full for us at this point.

Adrian Cohn: Part of managing any program is knowing when you've achieved enough. You don't need to go over. Jordan, this has been a really fun conversation. I feel like I've learned a ton and I have to say, I think you have done some pretty amazing work with your team at Culture Amp to realize an automated translation solution. Well done to you and to your team.

Jordan Lewis: Thank you. Thank you very much.

Adrian Cohn: And it's been a pleasure having you on the show as well. Be well, and look forward to seeing what you do next.

Jordan Lewis: Cheers. Thank you.

Adrian Cohn: I hope you enjoyed that conversation with Jordan. As much as I did, if you liked this episode of The Loc Show, hit the subscribe button. So the next episode we'll be waiting for you. And if you loved the podcast, please leave a review. Five star reviews go along way. If you're not ready to give a five star review, give our next episode a shot. We appreciate your listening. If you have any feedback or wants us to interview one of your favorite people in localization, just email me at locshow@smartling.com. See you next time.