Audio journaling & wellness insights app powered by Natural Language Processing (NLP).
Name: Monolog: Wellness Insights
Role: Experience design, visual design, iOS mobile design
Timeline: April 2020 - Present
What is Monolog?
Monolog is an audio journaling and wellness insights app powered by Natural Language Processing (NLP). Our goal is to empower individuals to understand the why in their emotional fluctuations. We know it’s OK to have ebbs and flows in your mood, but who or what causes those changes in your mood?— and better yet, what are we going to do to act on that knowledge in the future? Download Monolog HERE!
I joined Monolog in April 2020 as the founding product designer. As the only designer on the team, I was responsible for identifying user needs based on current and previous research, redesigning the iOS app interface, creating a new visual identity, and iterating my design based on feedback from users and developers.
Our user research found that 60% of people have tried journaling, but only 15% of this group can satisfactorily maintain their journaling practice.
Solution & Core Features
1. Audio Journaling Experience
Journal freely or take inspiration from our reflection prompts. Not enough time to write it out? No worries, all you have to do is think it then log it by speaking.
2. Wellness Insights
Through each journal entry, Monolog offers analytical breakdowns of your thoughts. Using natural language processing and artificial intelligence, there’s data that allow users to see trends and patterns in what occupies their mental space.
3. Curated Content
In addition the analytics provided, Monolog also uses your responses to provide you with curated, educational content, along with recommendations on how to achieved your goals. These resources are unique to you and your experiences.
We talked to potential users and current users of our beta (consisting of people who have experience with journaling and people who are new to it) to learn more about their current journaling habits and what problems exist in their routine.
General findings about journaling habits
- Life under COVID has been a time of introspection — many people have turned to journaling in a time of uncertainty.
- 60% of people have attempted to incorporate a journaling habit into their wellness routine but only 15% could maintain this habit.
- They often feel stressed but are too busy and always on-the-go.
- Most people currently use a mix of digital and physical journaling.
Why do people journal?
- To unload their thoughts and emotions, and learn more about themselves throughout the process
- To reflect on specific events / topics
- For record-keeping
Where do people usually journal?
- Physical journals
- Notes app
- Voice memo
- Other text-journaling apps (e.g. Reflectly)
Problems of traditional journaling
- Inconvenience: Many users felt like they don't have enough time each day to sit down and write. Physical journals are also inconvenient to carry around and prone to loss.
- Lack of a centralized platform: Journal entries are scattered across many different apps and platforms, which makes it difficult to keep track of entries and to see the relationship between entries.
- Lack of feedback: Some of our users expressed that it sometimes felt like they were journaling to the void, and were not seeing much improvements in their mental health.
After synthesizing the insights from user research, my team and I determined a list of important goals we wanted to achieve through this next version:
- Make it more convenient to journal anywhere anytime, by designing an intuitive and smooth experience for both audio and text journaling.
- Present users with relevant, interesting, and digestible insights on the trends in their wellness over a period of time and the potential reasons behind such fluctuations.
- Structure the app in a way that seamlessly bridges various functionalities, including the mood update, journaling experience, analytics, and content feed.
- Establish a consistent and approachable visual style throughout the app that aligns with the company's branding.
Existing Beta Version
I began by reviewing the existing beta version of the app, to identify areas that can be improved.
By communicating with the backend and machine learning engineers during the ideation stage, I was able to learn about the technical constraints in the early stage of my design process.
I learned that depending on how the user is speaking, the transcription can be occasionally inaccurate, which can lead to incorrect extractions and analysis (e.g. showing that you talked about something that you never mentioned).
After this discussion, we agreed to implement the following features:
- Daily Update: Engage in mood update + audio / text journaling for a flexible experience
- Reflect: Journal to a prompt chosen from our collection of questions to get inspirations.
- Edit entry: Edit the audio transcription and extracted topics if they are inaccurate. Users can also add tags to categorize their entries. This accounts for the transcription constraints.
- Explore page: View curated content (articles, quotes) related to mental health and wellness.
- Analytics: View NLP-powered visualizations and insights by metric: happiness, stress, anxiety, and focus.
Deep Dive — Journaling Experience
1. Establish goals based on user habits
From user interviews, I learned that many users struggle with the inflexibility of traditional journaling methods, and are currently using a mix of digital and written methods to satisfy both text and audio journaling.
With this in mind, we wanted to create a centralized journaling space that achieves:
- Easy transition between audio and text journaling, so that users can journal anywhere anytime with the method that they're most comfortable with.
- Flexibility of journaling freely about their day or with inspiration from prompts, through the options of Daily Updates and Reflect.
2. Map out journaling flow
I began by mapping out the user flow for a complete journaling experience.
3. Collaborate with engineering on animations
Based on results from user research, we learned that many of our users wish to make journaling a consistent habit because whenever they did journal, it was a therapeutic experience.
Therefore, going into designing the recording experience, we knew we wanted an immersive and responsive environment.
I started by doing some research online to get inspirations for potential sound wave animations.
Since I don't have any experience with animations, I asked our frontend team whether it would be possible to implement a sound wave animation that responds to the user's voice.
One of our engineers brought up that he previously built a small audio visualizer program. The waves can be adjusted based on bins, gain, and gravity to display different kinds of visuals.
I decided to use this visualizer program to create our sound wave animations. The goal was to find an appropriate setting that did not move too quickly or drastically, as that can be overwhelming to look at while journaling.
I spent some time playing with different settings of the program until I was satisfied with the visuals, before handing off the settings to frontend.
4. Iterate on recording screen design
While waiting for the animation, I iterated on the design for the audio recording screen.
I created the first version on the left as an initial attempt, and immediately identified issues:
- The white background works well with text journaling, but seems out of place with an audio recording experience.
- There is no distinguishing factor that signals the start of a journaling flow, which takes away from the immersive experience we wanted to achieve.
With these in mind, I created the second design using a gradient background to establish a full-screen, immersive recording experience.
The same background is used for our prompt-selection screen (below), to ensure consistency across parts of the complete journaling flow.
Final recording experience:
Deep Dive — Wellness Insights
1. Learn the constraints and goals of analytics
First, I discussed with our Machine Learning Engineers to learn about the types of insights that we are able to extract from entries, and the type of analytics that we can produce.
Then, I collaborated with the Product Manager to decide on which insights are most appropriate to present for each metric. Based on these learnings, I created rough wireframes of the types of graphs that best communicates the insights.
Following the rough sketches, I created mockups for each type of visualization:
2. Consider the edge cases (missing data)
During a check-in meeting with our engineering team, I was given the feedback that developers are confused about how the trend graph for each metric should behave in cases where we have missing data on certain days.
For example, what if the user completed the mood update on Monday and Wednesday, but not Tuesday? (see figure)
Based on user interviews, we learned that rarely do people consistently log mood data every single day — on average, people use the app 3-5 times / week. With this in mind, I began to ideate on possible ways of handling missing data in our graph:
- Using sections of dotted lines to represent days without data
- Using grey gradients underneath trend line
- Using blank space to block out specific days
- Using dots to represent each data point, connect with regular line
I presented the list potential mockups to our frontend engineers, and received feedback on which options are feasible, and which ones are not.
Based on frontend feedback, which eliminated some options, we decided on solution #4:
- Dots intuitively indicate the presence of data
- By using a continuous line connecting each data point, we are still able to present the trends and patterns of the user's mood fluctuations.
3. Organize the graphs
Due to the large number of visualizations we wanted to showcase, it was a challenge to display them in an intuitive way without overwhelming the user with a crowded interface.
I began by doing some research on existing data-driven apps in the health and fitness category, to study how analytics are organized. I then considered the following options:
A. Continuous scroll of cards
- Pro: Clean interface. Pressing on cards leads to a separate page containing the graphs.
- Con: Difficult to compare trends across metrics, since each metric is embedded in a separate page. The order in which we display the different metrics can imply a hierarchy of importance, which we do not want.
E.g. Apple Health
B. Toggle between graphs for each metric
- Pro: Can easily compare between metrics and graphs. Intuitive to see which visualization corresponds to which metric.
- Con: Requires a lot of toggle buttons and implied gestures, which can lead to a steeper learning curve for the user, and a more crowded interface.
To find out more about user preference, we conducted a series of interviews with our current users and asked them what kind of analytics-related functionalities they were most interested in. The main responses we received were:
- Ability to directly compare insights within / between metrics (e.g. If their happiness and focus levels are both increasing, how the context between happiness and stress differs)
- Ability to easily reference a graph when talking about it to their friends or family (e.g. "The happiness tone graph")
With this in mind, we decided on option B to organize the graphs under each metric. I included labelled toggles and descriptions to clearly explain the purpose of each visualization.
Prototype of final design:
We successfully launched our app on the App Store in August 2021. Download HERE!
By the numbers
~280 Invited Users active on the app!
Through post-launch interviews and the feedback form within the app, my team and I received many personal messages expressing how much they enjoy using monolog.
Here are some testimonials:
- “It was easy to understand. Also the visual style is soft and clean”
- “Love the app! I've been sharing it with all my friends"
- "I already journal consistently with Monolog. I love the instant feedback of extracted subjects."
- "I'm super fascinated in data and information, so I really like the feature of being able to see trends in my mood changes"
This was the first time that I designed an entire app from start to finish. Through the challenges I faced, I learned a ton of valuable lessons. Some of the key challenges and learnings include:
- Making significant changes based on user feedback and engineering constraints
- Considering the many potential edge cases
- Learning the best practices of information design, which is an area I'm new to
- Collaborating effectively with PM, engineers, and content
We are currently working on an updated version of the app to enhance functionalities and fix any bugs that came up after launch. Here are some updates in the works:
- Introducing "quick notes" for users to jot down thoughts without going through mood update
- Introducing a full onboarding process, including a walkthrough tutorial of the app's features
- Conduct further user research to understand the full journaling process for various personas
- Conduct user testing to identify areas of high and low usage
Since our first launch was invite-only and only open to a limited amount of users, we hope to have a full launch early next year open to all users. I am excited to continue designing and iterating our next best version of Monolog.