How Localization Supports New-Market Launches

Launching a product with confidence means knowing, not simply believing, you understand customers in a new market. When you truly understand how in-market customers will use, perceive and value a product, only then can you tap into reliable revenue streams in a new region.

Conversely, launches that fail to adequately validate the product in a new market are susceptible to many different means of failure. Poor translations, improper date formatting, regulatory crackdowns and culturally inappropriate designs can disrupt a new-market launch almost immediately.

So, why don’t companies put more emphasis on localization testing? Simply put, it’s an exceedingly difficult task to complete in house and at scale.

In a recent webinar called Localization Testing in the Real World, an Applause panel of experts provided advice for organizations that want to get the most out of the approach. Using examples from the news cycle, our experts explained how a simple slip of the tongue can result in hundreds of millions of dollars in lost revenue. Also, our experts referred to examples of real Applause customers who conducted successful launches with the help of real in-market crowdtesters who validated products throughout the software development lifecycle.

View the on-demand webinar here or continue reading excerpts from the conversation with our panel of Applause experts:

• Maria Mondragon, Solutions Consultant

• Phil Slocum, Solution Delivery Manager

• Jonathan MacLennan, Director of Enterprise Sales

This transcript has been edited for clarity.

David Carty: We're going to start with Maria. How are you doing today, Maria?

Maria Mondragon: I'm doing very well. Thank you.

Carty: So we have three key numbers here. Obviously, we have a little bit of a description of what these are. Can you tell me a little bit more about what these mean for localization testing?

Mondragon: Yes. So, here we have three numbers that actually relate to localization testing. The first number is 7,000 — that basically means the number of languages that we have around the world. This, of course, includes dialects in different regions. But, the question here is, how many does it really take?

In 2014, [it would take] 14 languages [to reach] 90% of the GDP. And the forecast for 2027 [says] you'll need 17 of those languages in order to cover that 90% of the GDP. So, you can see that with the low investment, you would actually get a higher return on investment, focusing on those main languages.

Then we have the 55, which basically means that 55% of consumers would actually rather buy from websites that portray everything in their own language. What this means is, customers will trust more those websites that will actually communicate to them in their own language.

Then we have the number 3, which relates to [Applause’s] localization tester requirements. Here we have (1) the fluency of the language. So, basically, being able to read, write, and speak in a specific language. (2) Understanding digital products — when you're talking about testing, you need to be tech savvy in order to be able to test these applications or websites. And (3) reside in region — why is this important? Because not only [do] we talk about translation, but also consider the cultural differences, and be able to speak to a specific audience in their own language, and also understand their culture.

Carty: Okay, that all makes sense, Maria. Now, I'm told you have a question for me on this slide. You asked me to put together these images of fruits, and I see some straws. I'm hoping we're making cocktails here, but you tell me.

Mondragon: We'll get the cocktails after the webinar, David. But, basically, what do you see there in terms of the green fruit? What is that for you?

Carty: Well, not to get ahead of myself, I feel pretty confident in this answer. The green fruit there, that's a lime, clearly. I would say the yellow fruits, those are lemons.

Mondragon: Actually, it depends who you ask. So, if we look at those two fruits, in Latin America, the green fruit will be called “limón.” But, if you ask someone from Spain, they will tell you that's actually a “lima,” so a lime, and vice versa. So depending on where you're at, the same object can have a completely different name.

We also have some straws. And if you go all over Latin America trying to find a straw, you'll have to make sure that you know at least three or four different names for a straw. So in Honduras, for example, we call it “pajilla.” In Mexico, they call it “popote,” and there's also other names, like pajita. All of these different names are relating to the same object. So, you see that depending on where you're at, a specific object can have different names.

More than language translation

Carty: So, I think we can admit that there's a lot of nuance to localization. And language is not the only example. Localization really touches on a lot of different areas.

Mondragon: So, when we talk about localization testing, this is a lot more than just translation. So we talked about earlier with the lime and lemon example, there are alternative meanings. Depending on the country where you're at, you can have the same object having different types of words. This is important to take into consideration when we're doing localization testing. You need testers that reside in country in order to know what the word is, and also know the meaning of that word in that specific region.

Then we see text truncation. Depending on what language you're translating to, the word can get a lot longer. So, if we talk about string length and how developers would make their applications, they have to make sure that when they translate a word, it will also fit longer strings. For example, in German, usually the words and phrases are a lot longer than in English, so you need to make sure that your application will fit all of those words in order to convey the right message.

If we talk about date formats, depending on the region, you would find the month first and the day later. So imagine in an e-commerce website, it's not the same thing [if] you'll get your objects the 12th of January or December 1st. That's a completely different date.

If we talk about colors and also images, this is also very important. Depending on the region where they're at, they might have different meanings. So, if we talk about the color red as an example, in the U.S. and in other parts of the world, this might mean danger or something bad. But, if you talk about red in Asia, this is usually celebratory, something happy, a color that is lucky.

We also have non-translated text. Usually in applications, you will see this a lot whenever you try to change from one language to the other: a lot of the words might be missing and they're not translated. That can cause a problem to consumers and your customers [who are trying to] understand what you're trying to tell them in that specific case.

Carty: [There are] other examples, including culturally inappropriate text or graphics, and that brings us to our next example here. There can also be real business consequences to landing on the wrong side of things, especially in today's geopolitical landscape. So can you tell us about a couple of these examples?

Mondragon: These were on the news. For [the] Fast and Furious [movie franchise], the actor John Cena was in an interview, and he said that he was really happy that the first country that actually watched the movie was Taiwan. What's problematic about this is that the Chinese government heard about this news. They didn't really like it, and what they did was they banned all of the franchise movies to be seen in China. So, you can imagine that's a real impact in revenue, because a lot of people in that specific market were not able to see any of those movies for a period of time.

There's also another example with Marriott where they listed in their website Taiwan as a country. [Consequently], within China, they banned their website for a period of time, and this is also a loss of revenue for Marriott.

Carty: It's important here to take a step back for a minute. We're all individuals, we all have different reactions to what's going on around the world, human rights issues, geopolitical issues, whatever the case may be. And at Applause, we're certainly sensitive to this fact, especially where we have testers all over the world.

But, simply put, if your company does business in any country, region or market, you need to understand how your product messaging is going to be received in that market. China is a huge market for Hollywood, as it is for many industries. In these cases, these mistakes had huge financial impacts.

Many organizations do try to conduct localization testing on their own. But, there are some real logistical challenges to trying to do this. So what are some of those issues that they face along the way?

Mondragon: With localization…. This is not only translation, but it conveys a lot of other things that you need to take into consideration.

One of the challenges that companies would have is the actual comprehension of the language, being able to translate and understand the meaning of a specific word — what we talked about in alternate meanings of a word — but also understand the culture. That's why you need people that reside in the region and that also know the country. If you can imagine having a person that speaks different languages, that doesn't mean that they also understand the culture of a specific region.

They also have in-house constraints, so the lack of bandwidth. You can imagine all of the testing that needs to be done for websites and applications. Sometimes they don't really have the bandwidth to go into a specific region and do all of that testing.

And out-of-context translations. We talked about what a word might mean in different regions. You need to make sure that you have the right words for those regions, and you also have to make sure that you are within the context that you want to communicate to your customers.

And biased translation. Usually when you have a translation company doing the translation for your application, you might not want them to be doing the QA or the testing for it. You need a third party and a fresh set of eyes that can go through your application from the outside and actually give you feedback of that application from an in-country and in-region tester.

Localization testing challenges

Carty: Thank you, Maria. OK. Let's welcome Phil to the party here. Hey there, Phil.

Phil Slocum: Hi, David. How are you?

Carty: I'm doing great. Thank you. So, your job involves sourcing experts for localization testing all around the world. Now, we've established that localization testing is really a complicated and challenging effort for many businesses to conduct, especially in-house. There's lots of room for confusion [and] missteps. In your experience, how do many businesses approach this task and where do they go wrong?

Slocum: Yeah, we see some customers who try to do this localization in-house. If they are an international company, they may try to find some employees in the correct region. But, often, they'll just ask a friend of a friend who may live in that region to do the testing. And, often, this results in a couple of problems. One, is it's very time consuming to source these people, because they're not dedicated testers — they're doing something else. So trying to coordinate that could take weeks, if not months, to get the testing done.

The second thing is, the bug reports that these friends, these employees, may provide may not be high enough quality. So, it causes a lot of extra time on the part of the product managers or the engineers to figure out what the issue is, and to be able to schedule it into their next sprint and get these things fixed.

Then, finally, it's always risky to only have one or two people testing an application. If you only have one or two people, there's often things that could be missed. So, it's always good to have a larger team that does localization testing on your product.

Carty: Let's get into some real-world examples here. I know you worked with an international technology company — we're going to anonymize these examples — and this company specializes in GPS technology. They were really struggling to provide a consistent in-car experience, messages were getting truncated, and it was really causing a problem in a new market expansion for this company. Can you explain what was going on there, and how we source testers on the ground to help solve that problem?

Slocum: They wanted to find out how their product worked on the ground in a car in Paris. The only problem was, they didn't have any employees in Paris. Their nearest branch was in Germany.

So, two people from the German branch got in a car and drove to Paris. At that point, they contacted one of our testers who lives in Paris who's native from that area. [That person] knew the language and all of the points of interest in Paris. He hopped in the car with them, and they drove around for about two hours and tested the navigation in that area. We were able to find issues related to translations, the most common routes to these points of interest — make sure that those were what was expected in the region, listen to the voice giving the directions and make sure that was all accurate. So, it really helped them a lot to vet this before final launch.

Carty: Sometimes you have to get creative and figure out creative solutions to interesting problems when it comes to localization.

Let's go into one more example here, if you don't mind, Phil. You recently worked with a delivery platform that was expanding to Germany. Now, that company ran into an issue with some German alphanumeric characters from a particular email provider, and that was affecting the customer onboarding experience. Can you explain what happened there, and how we helped resolve that issue?

Slocum: One of the more common email providers in Germany uses a character that's not a typical English character. So, when these users tried to create an account, which is the most important part of any product, they were blocked. Their email address was not accepted.

So, we were able to catch this bug in their soft launch, and they were able to fix it within a few days, so that by the time they had their full launch, this bug was fixed. So, obviously that could have had a pretty high-revenue impact to the company had [the defect] been alive.

Carty: Just one of those things that you can't even perceive of necessarily in a lab environment. So, that was a good catch for that company. Thank you, Phil.

How to approach localization testing

Carty: And now we're going to welcome back Maria. One of the places that organizations run into trouble is with validation. Can we explain how this occurs, and why double validation is a really helpful technique?

Mondragon: When we talk about localization testing and the results that you're getting, it's really important to validate those results. One of the methodologies to use is a double-blind validation. What this means is that, basically, you will have two testers work independently and come up with a specific word that they would use for a specific application. What this does is, depending on who you're talking to, you would select the specific word that will convey the message that the company wants to convey to their customers.

Just some examples here that you see is, for example, what does “wicked” mean to you, David?

Carty: Well, as a person born and bred in Massachusetts, “wicked” typically, as a regional term, means it is amplifying a word that's describing something. So, “wicked good,” “wicked far,” something like that. So, it's a little bit of a different definition than what we might experience around the country.

Mondragon: Exactly. So in other parts of the country, “wicked” might mean something bad. So, depending on what message you want to convey, you would choose one or the other word.

Then there's [another] example, in Spanish where you have colectivo. This word can mean a bus; in some places, it can be a group. So depending on what you want to say, that's the word that you would choose. That's why you need more than one option in order to choose [which word or phrase] is appropriate depending on the message that you want to convey.

Then, usually you need a neutral party that will be like a tiebreaker [to choose] between the words that you will have as the option that will best convey the message that the company wants to tell their customers.

Carty: Wicked interesting, Maria. So, when you perform localization testing in-house, that means you're often doing it in production. You've got a finished product that you're launching in a market to customers or prospective customers. Why isn't this the best approach?

Mondragon: Usually when you're testing in production, it's too late. So, once you're in production, your customers have already seen the negative impact, and this will be a bad impact for your brand and your brand reputation. If we talk about Marriott and John Cena, what happened to them [in China], it was a bad reputation in a specific region. You don't want those things to happen.

You need to be able to shift left and be able to test earlier within the SDLC in order to find these problems before they go into production. Once they go into production, it's usually too expensive to fix these problems. There's usually a delay from the time that you find the problem, and also from the time that you're able to fix it. And it can cause a revenue loss.

So, like the example that Phil was using [with email registration in Germany]. That was something that if they didn't fix as fast as possible, it was impacting their revenue directly. This will also give your customers less trust in your brand. Usually customers want to buy from brands that they trust. So, it's really important to be able to test before you even go to production.

Carty: Of course, and this is challenging, because more than ever, content is dynamic. In addition to product releases becoming more incremental, customers can also contribute on many platforms now, so that creates another whole dynamic to test with localization. All of this creates a scale problem for businesses, more launches, more markets, more content all the time. So, how should companies best scale localization testing?

Mondragon: So, as you scale and you have more dynamic content, there are certain things that you need to take into consideration. The first thing is to prioritize the markets. Whenever you're doing the testing, you need to make sure which markets you want to focus on first. It doesn't mean the other markets are not important, it just means that you need to focus on something first and then scale up to the other locations where you want to test.

You have to understand the launch strategy, the rollout plan, which countries do you want to go [into], and what regions. Also [what’s] the timeline for that, and when do you want your app or website to be available?

You also [have to] really define the remediation window. What does this mean? Whenever you're testing, you need to make sure you have time to fix any problems that you encounter before you go into production. This will allow you to not have the problems that we've seen before, like the examples that Phil mentioned and the examples that we mentioned at the beginning.

And always establish and maintain best practices. As soon as you have a plan of what you need to do, always keep in mind the best practices, and try to follow those for every market that you're launching in.

Localization testing examples

Carty: Thank you, Maria. To go back to one of your previous points, all these nuances, all these challenges to overcome, it all factors into your brand's reputation. First impressions matter. When you launch into a new market, it's not good enough to just simply hope that you're going to deliver a great experience for those customers. You need to be as sure as possible. Because if you're wrong, you might not get another chance at it. You might not get another chance to connect with those customers.

So, that's true for companies in all industries who are looking to launch into a new market as we'll learn from a few examples. Let's bring in Jon MacLennan here, our Director of Enterprise Sales. How are you doing Jon?

MacLennan: Hey, David. How are you?

Carty: I'm great. Thank you. So let's get into our first example here. This is a global employment website, and they were having issues with their open job listings. They weren't properly translating. This was leading to mismatched candidates applying for jobs. So tell us what happened there, and how we were able to quickly scale up a mix of both functional and localization testing to address that issue.

MacLennan: Sure. This customer was doing an overhaul on their website applications and mobile apps, and they were focused on localization across about a dozen or so markets. That being said, some of those markets were more important than others, and they really needed a flexible solution to focus more on those primary markets when needed while not abandoning the markets that were not as high of a priority.

So, [here at] Applause, we were able to align a dedicated and also dynamic team across their primary and secondary markets, which ran on two different sprint schedules, ultimately giving them the flexibility of picking and choosing where to focus their efforts on any given particular test run.

Carty: Adaptability, flexibility is so important when we're talking about managing different markets here. So, let's get into another example. This is another customer, an online travel company. We're talking about a household name here, available in more than 40 markets in 20 languages. This company had a localization testing vendor previously, but they had really struggled with them. And that had led to a lot of wasted time and a lot of frustration. So how did we dramatically reduce the defects that this company was seeing in production?

MacLennan: Sure. Like you mentioned, this particular customer, they were leveraging a vendor. The results really weren't meeting their expectations. Essentially one out of every three issues that were being found were either a non-issue or a duplicate. What that led to is internal employees were spending at times up to 40 hours a week combing through issues as the customer was live in about 40 markets at the time.

Applause is able to test their primary four platforms, returning results in typically around 24 to 36 hours, and we actually dropped their duplication and rejection rate all the way down to 1%. The real value add for this customer was the Applause delivery model and our white-glove service, as we have processes and people in place to triage issues and really reduce the amount of overhead they spent substantially.

Carty: It's a really interesting case there. One more customer example here, if you don't mind. This is a massive apparel company. Everybody knows this company. Now, as part of its market expansion, this company required localization testing in several dozen markets — massive scale. And they also had a need for some payment testing. How did we help this company out?

MacLennan: This particular customer had a primary focus on about 54 different countries. They were really looking for Applause to support them with in-market teams and ultimately across a very wide device/OS combination matrix. We were able to test seamlessly with their development process and ended up increasing their device coverage by around 360%. With so many markets to focus on, Applause ended up leveraging, at the end of the engagement, close to 1,200 people — of course, in their markets — which really met their needs from the device/OS perspective. The best part about that is that the matrix was truly representative of their customer base, because all of these testers were boots on the ground in those actual markets testing on real devices, which they found incredibly valuable.

Carty: Good luck trying to source 1,200 [people] in different markets. It's such an incredible task. It really speaks to that scale issue. So, thank you, Jon, appreciate it.

Emerging markets

Carty: Let's go back to Phil. I want to ask you about emerging markets. As we've established, there are countless examples of ways that you can deploy localization testing to create a higher quality product that better delivers customer experiences all around the world. But emerging markets really magnify some of these challenges that we're talking about. So what are some markets where companies are focusing new launches, and what are some of the unique considerations in these markets?

Slocum: We're seeing a lot of activity in Asia-Pacific, in Latin America, even some of the African countries. Some of the unique challenges for some of these countries is they may be very small countries. The [company] has very little visibility into how things work in those countries, from local regulations to the way a payment flow might happen in those countries.

So, in particular, when it comes to payments and finances, for instance, a user might be asked to link their bank account or their credit card to an application, and is that flow that the customer designed what the user is expecting? Especially from a financial viewpoint, if something doesn't look right to the user and it's going to mean money. If it's just a little bit off, there's a good chance they're going to abandon that payment. Obviously that creates a revenue loss for the customer.

So, it's really important to get that kind of feedback from these users even in the design phase. We can reach out to users in those small countries, like Costa Rica or Bolivia or Ecuador, and [ask], “What is the normal way that you add your credit card to an application?” And we can mediate those issues before they even happen.

Carty: That's why we say get closer to the customer. It's really understanding the customer's pain points. How the customer is going to interact with the product, not your own understanding of it, because you ultimately want to connect with that customer, with that person.

Slocum: Absolutely.

Carty: Thank you, Phil.

The Applause advantage

Carty: Let's go back to Maria one more time. How do we approach localization testing to help our clients get closer to their customers and realize their goals?

Mondragon: Yeah, David. The way we work at Applause is we have a global community of testers. We leverage that community in order to gather information about any type of testing that we can do. In the specific case of localization, we take into consideration the language and any cultural differences that they might have in order to get that feedback and be able to convey the best possible message to the consumers.

We also have a fully managed service. So, that means that we have a team that will take care and manage all of these testers. You can imagine for companies it's really hard to have different testers in different locations, that's something that Applause will take care of.

We also triage all of the feedback that is coming back. So, Jon mentioned this in his example. We get a lot of feedback, and we need to triage and give the most important information back to our customers. Now, our in-market testers, they will see all of the testing in a holistic way. That means that they won't only focus on translation; they will also see text that is not translated, text that is translated out of context, things that are culturally different and that might not be OK for a specific region, or even what Phil mentioned in terms of the payments — what payment methods are really common in those regions? Those are things that will help [companies] in order to give their consumers the best possible experience.

And we're an independent third party, meaning, this is a fresh set of eyes that will take a look at your application and give you actionable feedback on what you're portraying to your customers. [We’re] trying to give the best possible quality to all of the customers that you may have.

Carty: That global community that you mentioned is one million strong, and we've been working on cultivating and growing that community for years. It's not something you can do overnight. Not something that's easy to scale up in a new market when you're ready to launch, so that's why we are so enthusiastic about what we are able to offer our customers, and why I think that we deliver a lot of value to them as well.

So, let's get to the bottom line here, Maria. Companies have to grow, I'm stating the absolute obvious here, and one way that they can achieve that is expanding into new markets or new customer segments. What are some key takeaways that our audience should leave with today as they think about their own growth objectives?

Mondragon: I think the key takeaways from everything we've talked about, is aligning to your customers, how they think and how they operate. What this means is you have to make sure that you're communicating the right message to your customers, and you have to make sure that you speak their language. This will allow you to have trust with them, and they will come back to you in order to buy your products.

You also need to adapt as [you’re] moving to new markets and emerging markets. [You] need to adapt and see what is good for a specific market, and how you want to approach that market. If you don't do it, your competition will definitely do it.

And we have to think outside of our own wall. [Don’t] only think about our primary market. Once we know how our primary market works, we don't need to think that's the only way to go. We need to find out first how our customers in different regions will react to our messaging, and that way we can have better communication and have better quality for all of our digital products for our consumers.