Winning Localization Means Investment, Care, Testing, Trust

Nelson Mandela famously said "If you talk to a man in a language he understands, that goes to his head. If you talk to him in his language, that goes to his heart."

Maybe you’ve traveled to a country where you don’t speak the language. If you took the time to even learn a few words and then had the courage to use them on the locals, you probably noticed a wonderful thing about humans: we love it when people speak – or attempt to speak – our language. Even if your pronunciation was off and you simply said “thank you,” it likely helped create a bond, if ever so small.

However, when it comes to how companies communicate with customers and prospects, expectations are much higher.

Localization is incumbent on organizations

A 2020 CSA Research study of 8,700 consumers in 29 countries found that 76% prefer purchasing products with information in their own language. In addition, 40% will never buy from websites in other languages.

Today, global organizations and brands recognize that to gain the hearts and minds of customers in different markets around the world, they must make their digital properties available in multiple languages. This is table stakes. To do it right, to make your products and services really speak to your customers and prospects, there is no room for loose translation that misses the point, or worse, offends. And, it’s not simply about translation. Cultural considerations play heavily into representing your brand and products in various countries and languages.

This process of adapting a product or service to a different culture and language is called localization. When done thoughtfully and correctly, localized offerings create substantial expansion opportunities, improved UX, customer journey, long-term customer satisfaction and brand attachment. However, it requires forethought and planning that first enlists a blend of UX and market research, then functional and localization testing.

Core components of localization projects

I’ve been fortunate to have been working on global localization projects across all industries for years. In this blog, I wanted to cover a few characteristics of these projects and some insights as well.

Often, organizations are surprised to find that, despite their best efforts to localize on their own, elements escape their examination. Some can be critical. This is where localization testing comes in. We get into more detail around localization in this webinar, Localization in the Real World.

Here are a few basics around localization projects:

• Intention - Firms want to ensure that when they launch a digital product, app or a feature within a new market, that it actually speaks to their customers and that there are no errors in translation, no expressions that have alternate meanings and that proper dialect is used.

• Cultural context - Highly influenced by context, translations and related visual architecture must consider local customs and all cultural perspectives. For example, colors in one country may represent joy or fortune, while in another, may suggest caution. Graphics and other UI such as word length must be considered as some translated words will be much longer, impacting the size of buttons, navigation menus and general page layout.

• Dates - The U.S. and Europe abbreviate dates differently. In the former, the month comes first. This can cause confusion in countries where the day comes first. For example, in the U.S., December 1, 2023 is abbreviated as 12/1/23. In Europe, the date would be abbreviated 1/12/23.

• Currency and numbers - Firms must customize, for example, checkout processes to reflect local currencies and how countries show numeric value. Germany would write two thousand five hundred as 2.500,00, whereas in the U.S., this would be written as 2,500.00.

A few industry examples of localization testing

Retail

A large global clothing brand, which has particularly strong brand characteristics that it needed to represent, used a language translation provider (LSP) to do the translation from English to German. The retailer asked Applause to do verification that the LSP’s translation was localized accurately. We found some mistranslated words, as it can be very difficult to represent the concepts behind the words in another language. Verifying in region helped the retailer get as close as possible to the intended meaning. This was particularly important because the organization’s brand was very targeted to a specific demographic of customers. In addition, but unrelated to this localization translation project, when the company launched in Germany, it found that it was not seeing the conversions it expected. All functionality worked as intended, so Applause was enlisted to test localization and provide feedback.

B2B software

We support a global leader in B2B software which needs help doing functional and localization testing in 11 different languages, including Korean, Simplified Chinese, Traditional Chinese and Russian. In addition to requiring global coverage across a large number of devices and OS combinations, the company also needed a repeatable approach to in-country testing that could integrate in its Jira system. We do a blend of scripted and exploratory testing and find a wide variety of localization issues including: incorrect translation/mixed translations, untranslated text, text truncation, layout/UI issues, inconsistent terminology, locale issues (date, number, currency etc.) and Do Not Translate text issues.

GPS technology

Applause worked to help a GPS firm that was having issues providing a consistent in-car experience. It wanted to find out how its product worked in a car in Paris, but didn’t have employees there. Their nearest office was in Germany, and while the company could have used one of our testers in Paris without any accompaniment, they opted to have two employees drive to Paris. Upon arrival, they connected with an Applause tester, a Paris native who knew the language and the city very well. The Parisian drove the team around for several hours and tested the navigation, carefully listening to the voice giving the directions. The tester found significant issues related to translations for the most common routes to local points of interest. These issues were corrected before launch.

The anatomy of localization projects

Many of our typical testing projects involve functional testing, and we execute structured or exploratory test cases. Often, we do functional testing before we get involved with localization testing. For assessing functionality, we may use up to 15 testers. Through our years of experience, we’ve found this to be the optimal number of testers, as beyond that, we don’t tend to see any more bugs uncovered. For localization, however, we just use two testers per language project, selected for their strong translation skills and localization training.

Using two testers is optimal, as adding more tends to not yield more bugs. And, as they work separately, they arrive at their own suggested translations, graphic edits and more. They are uninfluenced by each other up to this point. This double-blind validation also means that if they arrive at different conclusions, they can challenge each other. We have a test engineer who ultimately mediates any disagreement and is a tiebreaker if needed, often reaching out to the client for buy-in in certain instances.

A tester may choose a word that is specific to a given region of a country, but not frequently used in other parts of the country. The company’s intention may be to find a word or phrase that can be used and understood across the entire country. For example, in Spanish, there are words that are specific to Spanish spoken in Spain and that spoken in Guatemala. The firm may be looking to have distinct sites for distinct countries, where that level of specificity is required, or it may be looking for Spanish that works well in all Spanish-speaking locations. Typically, however, market expansion projects are taken on by language and not countries. So, for French, for example, localization might be structured to work well in both Canada and France.

Applause is often involved with testing for market expansion for overall digital properties, but also targeted testing of feature changes within a specific market such as Latin America, for instance. In this case, we check with the client to see which country or countries they’d like to test in as best representative of the broader market. Customers may give us a guided exploratory scenario where testers are asked to go through specific web pages to search for localization errors or specific customer flows.

Localization maturity levels

Some companies have teams that work on localization. They are very competent, but just lack the geographic reach of testers in different locales. They will use an LSP and then ask them to do the testing. However, we don’t recommend this approach, as firms doing both translation and testing are prone to be blind to issues they’ve created.

Even when a company has done UX and market research and made a decision to launch, we often get involved in helping extend the UX insights because we have relevant experience in optimizing UX for specific languages/cultures.

Mature companies still make mistakes. Just recently, for example, I received a package addressed to me. My name, María, has an accent over the “í.” The package had my name written as “Ma?..?..a.” It was clear that their software couldn’t process the “í” with the accent. This wasn’t a major issue, but such oversights in a checkout process, for example, can totally derail a purchase.

Regardless of maturity, we always develop a scope of work that is customized to the needs of the customer. Localization testing can be a fully managed service where we recruit, vet, train and oversee all elements of a company’s project, or it can be a blend where we work as an extension of a company’s localization team. In either case, our testers capture details of all issues for the client: videos, screenshots, etc. We outline steps to trigger or recreate the bugs. Also, if we think there is an inaccurate translation, we always add in guidance on how to fix it.

Trust at the core

Like so many things in life, there are subtleties to how we communicate. A minor grammatical error in a communication can cause us to pause and wonder about the authenticity of the sender. There’s a spectrum of trust against which we consciously and unconsciously weigh communications. We have a gut feeling. At worst, a communication with errors signifies sloppiness. If they communicate like this, what kind of product do they produce? Somewhere in the middle of the spectrum, perhaps we think companies really just don’t care enough to check their correspondence. And, worst case, if a communication doesn’t clearly make its point, or even offends, then there’s little chance a customer or prospect is going to be convinced to engage more deeply with the brand.

Since most business interactions lack true personal relationships, it's key that organizations put the extra effort into ensuring that their digital properties work as intended, reach the broadest audience possible and show that they have taken the time and care to localize all content. The only surprises customers like are positive ones.