For the past few years, I’ve worked on sites and web apps that have large user groups of non-native speakers of English. That has given me a chance to look at how they are accepted (or rejected) by people who don’t speak English as a first language.
Some curious facts emerge when you compare the languages most sites use, versus the languages most internet users speak. While around half of all web pages are in English, only about 28 percent of the people using the internet speak English as a first language. Interesting, right? There are billions of people who use and browse the English web, but are not native speakers.
Asking for fully translated and localized sites is a mammoth task, one only large international conglomerates can afford. Instead, we can take some other simple steps to make our sites accessible for non-native speakers. We can focus on clear language, interfaces, and prompts, to help users as they navigate a largely English-speaking web.
A few years back, my coworkers and I were planning the design and launch of a marketing site for one of our new English language courses for adults. We started to gather copy and content. The majority of it described the features of the course and the resources available to teachers. We needed to keep in mind that many of the teachers for this course were not native speakers of English.
In order to get an objective idea of how complex our copy was, we ran it through the Flesch-Kincaid Readability Test. The test measures how difficult a passage in English is to understand, and assigns a score or US-reading level. You can do this with the “Show readability statistics” tool in Word, or by copy and pasting your content into an online tool. Our marketing copy was topping out around the 13th grade level, meaning you needed at least a year of college to understand what we were saying about our product! We should have been sticking to a level of 6th-8th grade to make sure our content was clear and accessible to our customers. No one wants to stumble through dense text just to get info on a new product!
In order to remedy this we needed to take a few basic steps, focusing on our copy, and our UI. First, we ran Flesch-Kincaid scores for all the content. We compared the passages that had lower scores with those that came in very high. Based on that, and what our marketing team needed, we decided to go for a reading level of around 9th-10th grade—clear, but not dumbed down for English teachers. We then created a content map, basically a spreadsheet with all the copy, its location in the site, the reading level, and a few other bits of information. For every place the readability was poor, we edited it down and ran it through the test again, until we were satisfied.
Standardized interface language
Another area where sites and web apps can trip up non-native speakers is in the interface language. Users with varying levels of comprehension employ all kinds of strategies to understand what they are reading. While there are many standard commands like “save” and “help,” other processes, like “upload image from computer” when written poorly, can be confusing for non-native speakers. We should aim to provide them with simple, standardized commands that they can easily read and remember, wherever they are on the site.
So how can we do this? I have found that creating a spreadsheet or list of all your interface language and sharing that with your design team is a great way to maintain a sitewide standard.
This strategy is not only relevant for non-native speakers. All users, regardless of their comprehension levels, need consistent messaging and instructions on your sites and web apps. By focusing first on those with the most pressing needs, you make your site usable for a much wider variety of users.
Sometimes though, simplified language is not enough. When we pair icons and text, we increase the speed at which recognition happens. It is an attempt to build a common language, one that is less dependent on English comprehension. But this is easier said than done. Look at these icons in Chinese mobile apps, and try to figure out how many you recognize.
Perhaps we need to offer an additional layer of context in these situations. Let’s start with onboarding. As non-native speakers download and enter your app or site for the first time, it is a great idea to provide small popups or a series of screens with animations to show the functionality. This handholding at the beginning of the experience can provide a sense of clarity and comfort even without a full understanding of the language you are using.
Tooltips next to key actions can also be helpful. At first glance you may think, “If it were designed correctly, you wouldn’t need tooltips!” which is true for native English speakers. But if, like me, your customers are not always native English speakers, then small, unobtrusive explanations next to key actions can provide that continued layer of help that your users need. Some web apps like LayerVault even use progressive reduction, where increased use means less explicit help is displayed, but if a user returns after a long period away, this contextual support is added back into the UI.
Non-native speakers of English often need a little extra help to get through English web interfaces. That is OK. If they are a significant part of your customer base, these are some simple ways to support them, making a more powerful online experience possible. Aligning your readability level with your users’ reading comprehension level, standardizing your interface, and expanding the range of help options available to users are all things you can do now—you don’t need to wait until the future when you have the time and money for a complete overhaul of your content. By planning and delivering these discrete steps, you can do a lot right now to help all your users, whether they are native speakers or not.
14 Reader Comments
I don’t think that Flesch-Kincaid has much to say about non-native speakers because it measures word and sentence length which affects native speakers as well and rather relates to the reader’s overall capacity to process text.
Judging from my own experience, understanding problems rather result from colloquial expressions, figures of speech or an extraordinary, literary vocabulary (and, of course, marketing blah-blah). I’d rather test how common the words in use are.
@Herbert: Word/syllable count and sentence length work well as a measure of readability for both native and non-native speakers though. Whether or not they were born to it, a user’s capacity to process what they are reading has a huge effect on the effectiveness of our sites. Best to tackle both user groups with an simple tool like Flesch-Kincaid, right?
Colloquialisms, slang, and other culturally dense language isn’t something I touched on here, but you make a good point. I think as a matter of course, we need to minimize our use of that language in our web/marketing copy. Even regional differences in English speaking populations can mean some head-scratching.
Good article on an under-discussed topic! I’m a UX specialist for a website based in the very multicultural city of Toronto, and have had many opportunities to do user testing with non-native English speakers.
I agree 100% about the importance of copy readability. In addition to evaluating reading level, another key copywriting strategy is to avoid overly-clever language. Wordplay that seems witty and original to your marketing department will probably befuddle your English-as-a-second-language users.
My other recommendation is to follow established design patterns. The search box needs to look like a search box. The main call-to-action should be strong and obvious. I’ve seen people with almost no knowledge of English successfully complete a task thanks to these familiar visual cues.
Interfaces aimed at a multilingual audience are not the place to push the design envelope and introduce unconventional language and radical visual treatments. This doesn’t mean your design can’t be sophisticated and attractive – simplicity can be beautiful too.
@Sandra: Good points about avoiding overly-clever copy and UI patterns. I’m guessing the more radical you go with your interface, the more the audience is just other designers!
Reading your comment, I wonder if research into the understanding/use of multi-lingual ballots would also be relevant here. That requires clear instructions and established interface patterns to be successful to a multi-lingual population.
A really interesting article, Senongo. Readability is a huge issue with the web these days. I’d go further than saying it’s about designing for non-native speakers, though – according to the Literary Trust 5.2 million adults in England can be described as functionally illiterate. That could be because of education, a developmental disability (e.g., autism, dyslexia), or something else. Even if a person has high levels of English language comprehension and is a native speaker, something like autism could mean they really struggle with metaphors or turns-of-phrase.
I’m finding that a lot of content authors are forgetting to write to their audience; most of the authors I talk to have at least a Bachelor’s degree and often slip into language that’s a far higher level than the audience’s.
That’s not to say that people misunderstanding something is because their language skills aren’t highly developed. In my last job I often found myself translating from web developer-speak into something non-developers could understand. This was usually because of an email, but sometimes it’d be in meetings where I could see that only two people around the table understood what the developer was saying but the people who had to agree to something had no idea what it was they were being asked to agree to.
The thing that I try to remember is that my audience isn’t me. My audience has to understand what it is I’m saying, they only receive the message. Keeping that in mind makes it easier for me to rework content without making it too simple; if I know who my audience are I can write it to their level. It doesn’t matter if they’re native or non-native speakers at that point; if they easily understand what I’m saying and learn something from it, then I’ve done my job as a content author.
I don’t feel that Flesch-Kincaid has much to say in regards to non-local speakers on the grounds that it gauges word and sentence length which influences local speakers also and rather identifies with the peruser’s general ability to process content.
@Karl Brown: Thanks for commenting! As you can tell from other comments here, issues of readability affect native speakers as well. My personal experience has focused a lot on non-native speakers, but I’d hope the lessons apply across the board.
As the writer, I would like to share with my opinion. Sometimes I meet sites with not grammatically clear content you need some time to understand “What did they want to say to me?”, “What should I do right now?”. There is no sense of making hard-understandable navigation. Totally agree, that site should be simple and standardized that users can easily read and remember.
P.S.: Really hard to pronounce for non-native speakers word like “Squirrel” 🙂
The technology is blooming each day and people have got new ways to connect to the Internet. Besides a laptop, now-a-days your website is seen on a mobile phone, a tablet, a phablet, and a smart phone. Thus, it is important for a website to be equally charming on smaller screens too. We offer responsive web design solutions that are efficient and can target numerous audiences through various devices – be it big or small.
Excellent read, wish there was more on this subject around. We have spent many an hour testing layouts and trying to build thai-friendly english designs based on english words that are more commonly used and easily pronouncable. Mixed results really, now we translate all designs (horribly time consuming) and find that reduces the local bounce rate on sites by about 10-15%.
Most of the people know English language than other languages and interested in learning this language because the English language is used in many websites and many companies are hired based on the communication. I have researched about this in many essay help and the studies shows that many website developers develop the websites in English language.
Good Article Senongo, Great topic to discuss about Non-Native Speakers. landscaping burlington
The language is the main tool to consider when you create a website. Because, the people will communicate any site by reading the website’s information. So, the developer must take English professional for the content of the website.
my last duchess
This strategy is not only relevant for non-native speakers. All users, regardless of their comprehension levels
Got something to say?
We have turned off comments, but you can see what folks had to say before we did so.
More from ALA
Personalization Pyramid: A Framework for Designing with User Data
Mobile-First CSS: Is It Time for a Rethink?
Designers, (Re)define Success First
Breaking Out of the Box
How to Sell UX Research with Two Simple Questions