I first researched competitors by reading app store reviews and online articles. I learned that users appreciate positive feedback, and dislike impractical content & the inability to correct mistakes.
One app store review for Duolingo wrote, “making mistakes is essential in language learning, we shouldn’t be punished for it.”
I signed up for popular language learning programs, then made a feature comparison chart. This helped me determine what features users have come to expect, like goals and feedback.
I then devised a survey which yielded 106 responses. I considered responses from people who have attempted language-learning outside of school, to ensure I was only considering data from target users. Many users quit was because they were forgetting what they learned or had lost motivation.
Next, I conducted 3 user interviews. One user "felt hopeless" when trying to learn a language. Another complained that online language learning was excessively repetitive (“it was ridiculous”). Another person said he was "barely able to communicate" when traveling, despite having made progress on Duolingo.
I also posted a question on reddit asking why people may have given up on a particular language-learning method. Most people said because it became too “boring” or “tedious." This research provided insight that the app should be fun and enjoyable to use.
I synthesized my findings in an affinity map to see patterns in the data. This also really helped me empathize with the users.
Using the data I gathered, I created a user persona that represents the target users. The biggest user needs are having an enjoyable and not overly repetitive way to learn Spanish, as well as retaining progress and learning about Spanish culture.
I then created a journey map to visualize the pains encountered throughout a language-learning session.
The top the pain points were:
1. Excessive repetition
2. Forgetting what they’d learned, and
3. Loss of motivation.
Next, I researched learning techniques and learned that spaced repetition is effective. With this method, users review difficult material more often than easy material. In testing, recall (used for fill-in-the-blank questions) requires more effort than answering multiple choice questions, but is a better indicator of learning.
I also researched learning and motivation. Interestingly, extrinsic motivation (points or grades) is easy to instill, but is short-lived. Intrinsic motivation (passion and genuine interest) is harder to instill, but leads to continued learning.
To guide my brainstorming, I devised “how might we?” statements for the 3 main user pain points (excessive repetition, forgetting what was learned, and loss of motivation). I like to use sticky notes during ideation; for some reason, this method yields better ideas.
I loved so many of my ideas but sadly, implementing all of them wouldn't have been feasible. In comes the MoSCoW method!
Considering impact on the user vs effort for the business, I determined which features would form the minimum viable product.
The minimum viable product is a mobile language-learning app for people interested in learning Spanish, that
• Allows users to rate session difficulty,
• Has few multiple-choice options, and
• Suggests music to complement lessons
The features solve the following pain points:
I hypothesize that this app will enable users to obtain longer-lasting knowledge, avoid frustrating & excessive repetition, and inspire motivation to continue language-learning. I'll know I'm right when users continue using the app long-term and the app is highly rated on the app-store.
The user flow helped me determine what screens I would need to devise lo-fi wireframes for.
I sketched a variety of lo-fi wireframes for each needed screen, then selected which ones to prototype and conduct usability testing for.
I conducted 7 usability tests to identify room for improvement in my design. Average task duration was 108.5 seconds with a 14.3% mis-click rate.
Users preferred a numerical scale to rate lesson difficulty, rather than simply choosing “easy” “medium” or “difficult.” I also learned that being given a suggested list of music after each lesson could interrupt learning and ultimately result in less lessons being completed. I made the needed changes in the mid-fidelity prototype, then did another round of usability testing before moving onto visual design determination.
I explored various options for the visual identity of the app. When conducting desirability testing of an early option, users described this style tile with words including "earthy," "money," and "brains."
I honestly cringe a bit looking back at this early version but it was part of my process! I identified a few brand attributes to better guide my design choices: modern, smart, fun, and cheerful.
To facilitate effective branding, I changed the app name to one different from that of the parent company (which is "Smart People"). I drew inspiration from online brand-name generators. I ultimately decided on "gluent," a play on words of "fluent," (which users will eventually become) and "glue," which represents the long-lasting nature of the learning.
I revised the colors and typography, and searched for icons with a consistent look. Users described the final style tile as being "fun," "clean," and "bright."
Finally, I arrived at the high-fidelity prototype.
In the future, I would explore the possibility of updating the music section to be called media, and adding Spanish TV shows and movies. The purpose would be to further cement learning, as well as to inspire cultural appreciation and thus intrinsic motivation to continue learning. I hypothesize that this would help minimize the amount of users who give up learning.
Success and failure metrics that would be used to determine product performance include amount of weekly active users, proportion of users accessing the curated playlist, rating on the app store, and churn rate.