Business

Web globalization lessons from Wikipedia

John Yunker

John Yunker

John Yunker

John Yunker

John Yunker is author of Think Outside the Country: A Guide to Going Global and Succeeding in the Translation Economy (available in both English and Japanese editions). He is cofounder of Byte Level Research and author of the annual Web Globalization Report Card.

Two decades ago, I began studying the globalization of websites. At the time, the practice was — like the internet itself — in its infancy and there were few established best practices. But by talking to the people who managed global websites and managing localization projects myself, I began to learn which practices worked and which did not. This led to the formation of an annual benchmarking report known as the Web Globalization Report Card. Now in its 16th edition, this ranking of the leading global websites highlights those websites that excel across a number of categories, including global reach, consistency and depth of localization.

This year, Wikipedia emerged number one overall. As one of the world’s most-popular websites, Wikipedia relies on the contributions of volunteer writers and editors from around the world in creating and translating billions of words across millions of articles. And while Wikipedia may stand alone in many ways, it exhibits a number of web globalization best practices that any organization can benefit from. There are five key lessons one can take away from Wikipedia’s global success.

The average number of languages on leading global websites

Figure 1: The average number of languages on leading global websites, 2005-2020. Source: Byte Level Research.

Lesson #1: When it comes to languages, there is no finish line

As I’ve said for many years, the internet connects devices, but language connects people. Companies that are serious about going global must be equally serious about supporting languages. Which languages you support will ultimately be determined by your global strategy, but if you were to follow the lead of the world’s global brands, you might aim for 30 languages, at a minimum.

Shown in Figure 1 is the average number of languages supported by the leading global websites according to the 2020 Web Globalization Report Card. Since 2004, this average has more than doubled to 33 languages today.

The leading languages of the internet.

Figure 2: The leading languages of the internet.
Source: Byte Level Research.

This average is based on studying the leading global brands, companies like Apple, Coca-Cola, BMW and Toyota. If we were to include a much larger sampling of websites, the average would most definitely drop. But one fact holds true — the steady, inevitable growth in languages across all websites in all industries. As more people around the world go online, the language requirements an organization must meet to communicate with them will only increase.

There was a time when English would connect you with the majority of internet users, but those days are long behind us. Figure 2 illustrates the leading languages spoken by the world’s roughly 4.5 billion internet users. Chinese is now the dominant online language, followed by English. But even if you support these two major languages, you’ll only reach 40% of all internet users.

An article in Russian and English that uses the same template

Figure 3: An article in Russian and English that uses the same template.

The largest slice of this pie chart is “all other languages.” This slice includes languages such as Italian, Dutch, Norwegian, Finnish, Swedish and many other languages that any global website probably already supports. What this means is that you can support the top 15 languages on the internet and still miss vast numbers of potential customers.

Support 48 or more languages, and your website will communicate with approximately 95% of the world’s internet users — which is the ultimate reason why companies such as Google and Facebook surpassed 48 languages years ago.

Which brings us to Wikipedia, with support for more than 280 languages.

Wikipedia is one of the world’s most multilingual websites, and a testament to just how many languages the world’s internet users want to see supported, as they themselves created this content. Yet even Wikipedia is far from complete in its linguistic journey; its language reach will extend as high as its user base wishes it to go. When executives ask me how many languages I think they should support, I often point to Wikipedia as evidence of what the world’s internet users want to see supported.

So even if you’re just starting out with a handful of languages today, prepare for a journey that will no doubt see this handful of languages expand over time. Web globalization is a road without end, and successful organizations understand how to design websites to support a diverse range of languages and scripts.

Lesson #2: Global templates enable efficient globalization

Your company may be only focused on expanding into one country today, but three years from now you may have 25 country websites to manage. In order to efficiently manage all of these websites, you need a global template.

Wikipedia could not successfully support so many languages across millions of articles without global templates. Figure 3 shows an article on the Pacific Madrone tree in English and Russian, both utilizing the same template. Should a volunteer editor wish to create this article in another language, this template will simply be reused.

Global templates allow regional and local teams to focus on content rather than design, a much better use of their time. And global templates provide a more consistent user experience across your many geographic websites and brands. Visitors may arrive at your global .com site and then navigate to a local website; if the designs change significantly, this may create an unneeded distraction.

Global templates may also be successfully leveraged across multiple brands. Figure 4 shows three Microsoft products — Microsoft, Windows and Office — all sharing essentially the same template. Given the benefits to global templates, one may wonder why all companies don’t support them. The major reason is that these companies are often quite decentralized, with independent regional and local offices that are not only comfortable maintaining independent websites but insistent on remaining that way. For those in the global office, it can be difficult to convince local offices to embrace global templates, but the effort is well worth it. Global templates allow organizations to be globally consistent while remaining locally flexible.

Figure 4 : Three Microsoft products share the same template.

Figure 4 : Three Microsoft products share the same template.

Lesson #3: Think mobile when thinking global

Most of the world’s internet users access the internet through their mobile devices. So it’s critical that the global templates you develop not only adapt to mobile devices, but are optimized for mobile users. The Wikipedia web page easily adapts its content to mobile and desktop browsers.

Adapting to the device is step one. Adapting to the user and their mobile needs and requirements is much more important. For instance, if the person understands only Russian, you don’t want to send English-only content.

Fortunately, Wikipedia uses language negotiation (also known as language detection) to greet users with the language most closely aligned with the user’s web browser setting. Language detection is one of four elements that go into forming an effective global gateway strategy, which also includes country codes, geolocation and a visual global gateway menu. These elements vary by organization but all can play a critical role in helping visitors quickly and seamlessly find their local content.

And what about offering users a mobile app? Wikipedia offers one as well, which features an interface localized into more than 50 languages, far more languages than most other mobile apps currently available.

What’s interesting to note about the app is how it makes navigating across languages within a given article easy to do. By clicking the languages button circled in red (Figure 5), a user can switch to a different language, which is a nice feature for multilingual users in search of all available content.

Figure 5: Navigating between languages is easy in the Wikipedia app version.

Figure 5: Navigating between languages is easy in the Wikipedia app version.

Lesson #4: Keep it “light” to improve the user experience

As companies expand their global reach, they inevitably discover that not all mobile subscribers enjoy unlimited data plans, let alone the buying power to pay for unlimited data. They also discover that many regions around the world are still dominated by slower networks and older devices. Which means that those companies that localize their websites to remain fast-loading and functional for users with older devices and slower connections are going to be perceived as more user friendly within these markets.

For a typical mobile web user in India, a web page that “weighs” 8 megabytes (MB) a page could take 20 seconds or longer to display. I mention 8 MBs because, according to the 2020 Web Globalization Report Card, the average weight of a mobile home page has more than doubled from 3.1 MBs in 2016 to 8.3 MBs today. This is a troubling trend.

Except for Wikipedia.

While most websites reviewed for the Report Card weigh well in excess of 1 MB, Wikipedia is one of the few websites to weigh less than 1 MB. The Wikipedia home page weighs less than the home pages for Google and Facebook, two companies obsessed with performance. One way that Wikipedia keeps its website so lightweight is the absence of scripts that track users and usage, something no corporate website would think of going without.

I recommend that all websites target a ceiling of 1 MB — an aggressive goal to be sure, but one that will ultimately benefit your customers and keep you ahead of competitors. Limitations can be highly positive because they force everyone to do more with less.

Lesson #5: In the end, it comes down to relevant content

Like so many corporate websites, the majority of content on Wikipedia is available in English — more than six million articles. German comes in second with support for more than 2.4 million articles, followed closely by French. But when it comes to less-used languages such as Gaeilge or Kurdish, you’ll find 10,000 or so articles, or fewer.

But just because you don’t offer parity with English on your website doesn’t mean you can’t create a valuable experience for visitors. As shown in Figure 6, Wikipedia highlights links to timely and newsworthy content.

Figure 6: Wikipedia’s Catalan homepage.

Figure 6: Wikipedia’s Catalan homepage.

This lesson carries over to all organizations. Just because you don’t offer all content across all languages doesn’t mean you still can’t offer useful and relevant content to users. But you need to have a plan in place for managing language and content expectations. By all means, do not make the mistake that many companies make by launching what I call “local façades,” in which only the home page and a few subpages are translated. Local façades create negative first impressions; visitors quickly discover how little is available to them and are reluctant to return, which could do far more harm than good in the long run.

And Wikipedia is well aware of the content and linguistic gaps. It recently acknowledged that it needs to invest in helping users in emerging markets such as the “Global South” to more easily create and edit content using mobile phones on limited-bandwidth networks. In the end, content is what will make your website, and any website, worth visiting.

Wikipedia remains one of the world’s most worthwhile places to visit. I find I land there a few times a week, and it’s almost always time well spent.

So the next time you visit Wikipedia, think about what amazing achievements have been made by soliciting the input of the world’s internet users. These days, there seems to be much to dislike about the internet, from privacy issues to online scams, but Wikipedia still remains, in my view, one of the shining accomplishments of this still-emerging internet.