How it started vs. how it’s going: a fresh look at DeepL’s accessibility efforts

author_by DeepL Team
Blog header image featuring a stylized version of the DeepL UI with a high contrast white on black background color scheme, with a green translate button and yellow highlights for navigation buttons.
contents_table

DeepL’s commitment to accessibility 

Accessibility is an integral part of how we develop and test products at DeepL. This year, we are celebrating Global Accessibility Awareness Day on May 18 by looking back at our ongoing accessibility improvements and achievements at DeepL since our update last year, starting with our Accessibility Statement. While improving accessibility is a long-term process, we’ve made some significant strides, and our improvements will continue as our product offerings grow and mature. 

A holistic look at accessibility 

Accessibility is not only limited to how users interact with the final product or feature e.g. DeepL Translator for desktop, or our Android mobile app, but also in how and with whom we conduct testing. 

This past year, we continued to diversify our testing pool, working closely with visually impaired testers who use screen magnification technologies, alternative color schemes (this allows us to comply with certain operating systems, like Windows, where users can select between multiple themes with higher contrasts), and various screen readers. Furthermore, we made sure to include testers who were dependent on alternative navigation methods, like voice control software. We used the feedback from these tests to help improve readability and control when using DeepL products. 

Screenshot of the DeepL translator page in a high contrast colour scheme with black text on white background. Screenshot of the DeepL translator page in a high contrast colour scheme with white text on black background and yellow and green text in key areas.

Consistent testing  

In addition to adding to our diverse tester pool, we’ve maintained a regular frequency of our tests. Since April 2022, we’ve conducted five tests per month, and we continue to use Fable Engage as our user testing platform. We highlighted this platform in our 2022 accessibility blog post, but we feel it’s worth reiterating what it does as well as the impact that it has on our users:  

Fable Engage enables teams to connect with people with disabilities remotely and on-demand to accelerate user research, design, and development. Feedback from accessibility testers provides invaluable information for companies looking to identify problems and to improve accessibility. 

Fable rates digital products using the Accessible Usability Scale (AUS). The score is calculated through a series of questions answered by accessibility testers, who assess sites on compatibility with accessibility technologies, complexity of site navigation, and ease of use without outside assistance. The AUS uses a 1–100 scale, with 100 as the top rating. 

Last year, our scores trended at 90 or higher. While the overall score fluctuates depending on our test results, we are proud to say that our score typically hovers around the same number. That said, when our score drops, it is a clear indicator that we still have room for improvement when it comes to both accessibility and usability. 

Getting it done: how we resolve accessibility issues at DeepL  

What happens when accessibility issues are flagged to us? Our team treats them as bugs, and a ticket is created for each issue. It’s our goal to resolve these as quickly as possible, and we are proud to say that in the last 300 days, our engineering team solved 63 tickets related to accessibility.  

Working closely with several teams at DeepL, particularly with those who help develop and improve customer-facing products, we make accessibility a top priority, ensuring that we keep it top-of-mind for all teams at our organization. 

Graph plotting the number of created and resolved tickets related to accessibility over the last 300 days. In that time frame, 75 tickets were created and 59 were resolved.Pie chart showing resolved issues per development team. Most tickets were resolved by the teams responsible for the website and native applications. Overall, more than 10 different teams are mentioned in the chart.

Praise for DeepL from our testers 

Although we are always aiming to improve our products, we also want to recognize the efforts of our team and highlight positive feedback from our testers. Here are a few quotes from a few people who enjoyed using our language selection tool when translating:  

"The DeepL Translator website is excellent to use and navigate around to locate different languages." 

"Truly wonderful! I absolutely enjoyed this request! I had no accessibility problems or issues." 

"I also found it to be enjoyable that you can also hear the translation if you so desire. This is so cool." 

"Since this is very accessible, this definitely gives blind people more independence. Once again, wonderful!" 

The full picture on testing  

At the time of the last blog post (April 2022), we were only testing the desktop web translator. Now, we’ve extended our testing to all other products to include: 

  • Mobile web page 
  • Windows and MacOS desktop apps 
  • iOS and Android mobile apps 
  • Browser extensions 
  • DeepL Help Center 
  • DeepL Write 

Not all are as accessible as the web app, but we continue to improve every year. 

It’s a marathon, not a sprint 

Accessibility is an integral part of our product development process and a quality criteria throughout all our teams. We are proud of how far we’ve come but promise to prioritize accessibility for all users as we move forward.  

If you ever have any feedback on our products, we’d love to hear from you. Thank you so much for continuing to use and help us improve our products and features each year.  

shareMenu_headline