Does anyone have actual usability study / testing research around the most effective solutions for language switches, or even dealing with linguistically fragmented content on an information website?

I'm working on an information website that's mostly in English, but has some pages in Spanish. I'd like to examine any established secondary research that shows what has been tested with regard to efficacy of switching languages, or perceived gaps in language content parity.

ux.stackexchange has plenty of opinions and commentary on the subject, but I've struggled to find actual usability testing results on the subject.

  • If you want to find a good article with research findings, you have to combine this search equerry with the word "accessibility", because if you want to create a proper solution, you also have to follow the rules of accessibility itself. The biggest issue with the language switchers, that they usually think about it as an aesthetics problem (how it will fit into the UI), and not as an accessibility issue. Commented Apr 13, 2022 at 7:25

3 Answers 3


I recall that Nielsen Norman Group published 6 Tips for Improving Language Switchers on Ecommerce Sites based on a study they did.

The give the following tips for language switchers in general:

  1. Default to the language version that matches users’ browser settings.
  2. On desktop sites, put the language switcher in the top corners (left or right).
  3. On mobile sites, put the language switcher above the fold or inside the navigation menu.
  4. Display the name of an available language in that language (e.g., Español instead of Spanish).
  5. Use a combination of symbols to represent the switch (but not country flags)

Unfortunately it does not cover the issue of fractional translations.

How we improved translations on GOV.UK explains that from a more technical perspective, in that exact context where only certain articles were translated, but not the surrounding menus etc. but there is no user research background. Their design system team unfortunately didn’t write on the subject.

To NN/g first tip I’d like to add that language should be dependent on the URL, so that search engines can direct users to the article in the language they picked. One good pattern is to add a banner if the browser language does not match the content language, offering the content in that language. After all, the user might understand that language perfectly fine.


This seems to be a very niche and specific question that may not have any detailed user testing associated with it.

When you are talking about the efficacy of language switches, do you mean whether there's loss of information in user understanding or if it disrupts the flow of information comprehension for the user? This would not be the case for a bilingual user for example, or if there are other features integrated with the website (e.g. Google Translate extension) that allows some level of translation of the content. Generally people who design websites will do so for a primary language and provide localization for the content.

However, if it is intended to be displayed in more than one language (with the majority in just one language) then you would expect it to be catered for because it is only a small fraction of the content.


It all depense from case to case, currently I work on a big SaaS that has option of creating digital qr menus. And they need that every option or product added to the menu can be translated. Google Translation was not the solution, so I just created different page and took Wordpress/Shopify route where you have few Tabs with sections to translate all of the parts. Something like this https://shopidevs.com/wp-content/uploads/2021/06/shopify-multi-language-online-store.jpg

  • 1
    While this is interesting, it does not answer the questions for some research.
    – Andy
    Commented Aug 8, 2023 at 7:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.