top of page

Case Study - Product Improvement: Duolingo


Duolingo logo


Background

Duolingo is the most popular language-learning platform and the most downloaded education app in the world, with more than 300 million users. The company's mission is to make education free, fun, and accessible to all. Duolingo is designed to feel like a game and scientifically proven to be effective. In addition to its core platform, the company created the Duolingo English Test, an affordable and convenient language certification option that is accepted by over 2,000 universities. The company uses the freemium model; the app and the website are accessible without charge, although Duolingo also offers premium services for a fee.


User Persona

An image of a table that defines the user personas for Duolingo

A typical Duolingo User Journey


Michael installs the app from the Android/Apple store and creates an account. He chooses the language he intends to learn and chooses the motive, level of preliminary understanding, and daily goal. He then goes through a 5-minute computer adaptative placement test that determines his understanding of the language accurately. His learning starts from there. Every day, thereafter, Duolingo sends him reminders, learning from his behavior, at a particular time of the day when he is most likely to act (learn).


Each step in this journey is powered using Artificial Intelligence, Read here:


Business Goals


Increase user engagement and enhance the learning experience on Duolingo App to lead to increased adoption of premium plans.


Customer Problems/Wants

  1. The user faces a problem in learning/understanding the native touch (in terms of customs and cultures) in language.

  2. The user fails to learn voice communication and has limited learning for conversational comprehension.

  3. The user wants to practice how to communicate based on a particular scenario - like ordering dinner but does not have engaging options available.

Ideas

  • Find Your Duo - Collaborative Learning among users. For example, a native Spanish speaker and an English speaker get together at a specified time of the day and practice together.

Explanation: With this feature, the users will have the availability of native speakers at their convenience to learn the nuisances of language. This will also enhance human-to-human interaction without physical presence. A sense of community will be built, and the user will enjoy the process of learning and communicating, along with learning how to speak like a native.

  • Audio Comprehension - Testing the understanding through an audio clip (Short sentences to longer as per the progress level)

Explanation: With current features, while the users are able to learn words and sentences, the comprehension in terms of conversational audio (which may be faster or slower in the case of native speakers, like native Spanish people speak pretty fast) is low. This feature will allow the users to practice with a native tone of the conversation and enhance comprehension in a real-life setting.

  • Conversational Learning - Avatar-based video conversation between a Duolingo Avatar (with the ability to choose from many) and the User’s Avatar.

Explanation: The scope of learning how to effectively communicate is limited to Duolingo. This feature will make it effective and fun to learn. An avatar of user choice (Say, superheroes, for example, Hulk - Green - Duolingo) and a list of fun and everyday scenarios will get users hooked to the app and enhance their learning curve. The User will also have the option to modify/design his own avatar as he wishes.


User Stories


Epic 1: Collaborative Learning


User Stories:

  1. As a user, I want to be able to choose the time and date of learning so that I can learn at my own convenience.

  2. As a user, I want to be able to see the details of suggested partners - such as their name, location, and profile rating so that I can make an informed choice.

  3. As a user, I want to be able to select a partner from the available list so that I can choose a learning partner of my choice.

  4. As a user, I want to be able to be matched to another partner with a similar rating score for learning in case the first partner cancels so that I do not lose out on learning.

  5. As a user, I want to be able to disconnect and block a partner so that I can maintain my privacy and user experience on the app in case the partner’s behavior is not appropriate


Metrics:

  1. Percentage of users who opted for collaborative learning

  2. The ratio of users who completed collaborative learning to users who opted for collaborative learning

  3. The average number of times a user completes collaborative learning

  4. The average duration of collaborative learning sessions

  5. Pre-feature average duration vs. post-feature average duration

  6. The language combination that saw maximum collaborative learning

  7. The ratio of users who blocked or disconnected from a partner vs. the of users who started collaborative learning

  8. No. of users who reported issues in using or accessing the feature


Epic 2: Audio Comprehension


User Stories:

  1. As a user, I want to have an option to practice audio comprehension so that I can improve my listening skills.

  2. As a user, I want to be able to hear, play, pause, and rewind the audio so that I can listen and answer appropriately.

  3. As a user, I want to be able to submit answers using text and audio so that I can learn to converse more effectively.

  4. As a user, I want to see if my answer was correct or incorrect so that I can put the required efforts to improve accordingly


Metrics:

  1. Percentage of users who practiced audio comprehension

  2. The average number of times a user practiced audio comprehension in a week

  3. The average duration of learning using audio comprehension

  4. Pre-feature average duration vs. post-feature average duration

  5. No. of users who reported issues in using or accessing the feature



Epic 3: Conversational Learning


User Stories:


  1. As a user, I want to be able to create my avatar so that I can personalize my experience.

  2. As a user, I want to be able to see a list of Partner avatars so that I can choose one of my choices.

  3. As a user, I want to be able to view, play, pause, and rewind the video conversation so that I can understand and participate in the conversation.

  4. As a user, I want to be able to record my answers easily so that I can participate in the conversation.

  5. As a user, I want to be able to see my conversational correctness score so that I can plan to improve accordingly.


Metrics:

  1. Percentage of users who created their avatars

  2. Percentage of users who selected a non-default partner avatar

  3. Avatar which was chosen the highest number of times

  4. Percentage of users who completed avatar conversation till the end

  5. The average duration of conversation completed

  6. Pre-feature average duration vs. post-feature average duration

  7. The average number of times a user opted for video conversation in a week

  8. Correctness score vs. frequency of usage

  9. No. of users who reported issues in using or accessing the feature

Wireframes


Feature 1 - Collaborative Learning with Native Speakers


wireframes for Duolingo product improvement suggestions - part 1


Feature 2 - Audio Comprehension


wireframes for Duolingo product improvement suggestions - part 2


Feature 3 - Conversational Learning Using Avatars


wireframes for Duolingo product improvement suggestions - part 3


Note:


In real-world scenarios, most of the time, it's not possible to develop all the features at the same time, and hence, we need to prioritize keeping in mind the Desirability, Feasibility, and Viability of the feature (Business Benefits, User Benefits, and Cost). Even in PM interviews, it's expected to delve deeper into a single feature that appears most rational.


Subscribe to receive updates directly in your mailbox.

Connect on LinkedIn to discuss more.

Comments


bottom of page