Language Learning on Duolingo — Now With Tangible Progress!

When people think of Duolingo, the first thing that comes to mind is usually Duo, the iconic owl. Say what you will about Duo, but his primary function is to accompany and encourage Duolingo users as they learn new languages. He’ll reward you with a streak for every day you learn your chosen language and cry if you quit a lesson without completing it.

Like I said, encouraging.

But users often give up on learning with Duolingo, even with Duo cheering them on. For all the app’s remarkable uniqueness, it clearly has a user-retention issue. I set out to understand where this issue is rooted and how design-based solutions could solve it.

How Could Users Abandon Duo?

I went hunting for Duolingo users to interview about this issue and immediately ran into people who used it for some language but gave up on it. They were the perfect people to interview given my research goals, which were to understand

1. Why people learn new languages and how it correlates to Duolingo use
2. If and how people stay motivated in language learning, especially on Duolingo
3. How people gauge their progress in language learning
4. How Duolingo compares to other language learning methods

As I conducted my interviews, I tried to figure out exactly how users went from excited and bright-eyed to unmotivated and uninterested. I came away with four key points:

I also found that users had a lot of strong feelings about Duo the owl. One even shared with me some fantastic meme content:

Credits to Ana for these 🥰

Aside from the mega-terrifying owl issue, though, here’s what I learned: Duolingo users want to learn a new language to a certain level of proficiency, but they can’t do this well because

1. They do not see their growth in the language over time
2. They lose motivation and get too bored to continue learning the language

Brainstorming: Pandemic Edition

Now that I had a problem, I just needed to figure out how to solve it. Should be pretty simple, right?

Thank you Max and Neha 🥺

I recruited two fellow Cornellians, and with the help of Zoom and a digital whiteboard, we brainstormed ways to Fix Duolingo™.

We found that our most effective ideas fell into one of two opportunity areas:

1. Making learning more practical
2. Improving the user interface

I focused on the first opportunity area because it was the most promising for introducing a new feature. After studying our endless digital sticky notes, I decided our best idea was the Non-Quiz Lesson, where users tap through a formal language lesson before taking the lesson quiz:

I moved forward with the idea, even writing out possible interactions with the feature and content requirements. Unfortunately for me, it turned out that I had forgotten one key point: my solution needed to be design-based. The non-quiz lessons were not that. They were more of an algorithm change, and that wasn’t my call to make.

Back I went to our digital drawing board, where I picked out a better, design-based idea: Progress Checks. With this feature, users receive updates on their progress as they make their way through the lesson quizzes.

I sketched the feature out to get a better grasp on how users would interact with it:

Now, I was ready to turn the Progress Checks into a fully-fledged feature.

If The Feature Fits…

I tapped and swiped and scrolled through Duolingo, looking for the best place to integrate the progress checks. The top bar on the home page — which I referred to as the “Language Bar” — made the most sense because it already housed other indications of progress.

Call Me Dora

The next step to integrating the progress checks into the Duolingo app was exploring different design structures. I created medium-fidelity mockups of ways to set up the feature.

In one of these explorations, the progress checks took the form of lesson summaries before a quiz and new skill descriptions after it.

With this setup, users could decide if a lesson’s content interests them before starting it, and once they finish it, they feel a sense of accomplishment knowing the skills they gained. However, the feature lacks direct encouragement to continue with another lesson.

I also wanted to try an approach outside of the quizzes themselves. In other explorations, I tried a dropdown to the language bar showing recently gained and upcoming skills.

One way to set up the dropdown

A different way to set up the dropdown

While this setup doesn’t offer specific connections between the skills and lessons, it does allow users to see recent and upcoming progress at a glance.

Designing in Feather Green

Armed with my medium-fidelity explorations, I was ready to bring my feature to Duolingo-worthy life.

I built a UI Design Kit with the help of Duolingo’s brand guidelines, carefully selecting my Feather Green and Bee Yellow and recreating the app’s complex icons:

I had everything I needed to create high-fidelity explorations of my feature (in dark mode of course).

Faced with the real visuals of the app, I had to make some changes to my previous design. For example, instead of creating a brand new skills dropdown in the language bar, I simply embedded the contents under the Crowns dropdown.

A hi-fi mockup involving the skills list

One of the interactions I added was a skills list, which users could access from the Language Bar dropdown. With this list, users would be able to see the skills they gained from each lesson as well as what lies ahead if they continue learning. Knowing when they’ll get to learn a skill they’re interested in, users will be more motivated to learn.

The Fairest Feature of Them All

With my presentable mockups complete, it was time for user testing. I wanted to understand if users felt they were making progress with the new features. If they did, I wanted to know which of the interactions was most representative of that progress.

And after conducting a few interviews, here were my key takeaways:

1. Users still associate Duolingo’s filling up circles with progress
2. They want access to both an at-a-glance summary and in-depth account of skills progress
3. Bite-sized progress updates during the quiz would further the sense of progress

So I went back to the drawing board, trying to figure out how to mash all my explorations into a single flow. I looked for ways to integrate those bite-sized progress updates into the lessons and include the lesson circles in progress summaries.

An exploration of mid-lesson updates

I settled on adding optional mid-lesson updates, which appear as nodes on the lesson’s progress bar. Interested users can click on these nodes to see what each point in the lesson means for their learning, and uninterested users can simply continue the lesson.

An exploration of integrating lesson circles into skill lists

I also settled on adding the lesson circles to the recent/upcoming skills window since users appreciated its at-a-glance nature. Adding a horizontal scroll of the skills (Middle 2.1) would have taken this away and made this piece of the feature more overwhelming.

Having done my best to address users’ feedback, I constructed a full prototype for my feature:

Conclusion

Duolingo is a unique language app. There really isn’t anything like it—I looked! No competitor offers quite the fun and supportive learning environment that Duo and his friends foster. But the key to learning a language is practice, so neither user nor Duolingo benefits from a loss of interest. Adding these progress checks would be the first of many necessary steps to retain users and catapult Duolingo to new levels of popularity.

I am a member of the Cornell University Class of 2024 pursuing a degree in Operations Research & Engineering. This was a case study of Duolingo I completed for Introduction to Digital Product Design. I am in no way affiliated with Duolingo.