A review that collects user insights for personalized audiobook recommendation
How to design an audiobook review that guides users to express their thoughts with ease?
A review that collects user insights for personalized audiobook recommendation
How to design an audiobook review that guides users to express their thoughts with ease?
A review that collects user insights for personalized audiobook recommendation
How to design an audiobook review that guides users to express their thoughts with ease?
ROLE
Product Designer
ENVIRONMENT
Remote
Tools
TIMELINE
10 weeks
TEAMS
CEO
CTO
PM
Developer
Tools
ROLE
Product Designer
ENVIRONMENT
Remote
Tools
TIMELINE
10 weeks
TEAMS
CEO
CTO
PM
Developer
Tools
ROLE
Product Designer
ENVIRONMENT
Remote
Tools
TIMELINE
10 weeks
TEAMS
CEO
CTO
PM
Developer
Tools
Overview
Overview
Booklet is an audiobook review platform that focuses on the narration aspect of the book and helping users to find their next audiobooks through the community.
Booklet is an audiobook review platform that focuses on the narration aspect of the book and helping users to find their next audiobooks through the community.
Goal
Goal
🎯
🎯
Design an audiobook review that collects critical user insights for personalized audiobook recommendation.
Design an audiobook review that collects critical user insights for personalized audiobook recommendation.
Problem
Problem
😥
😥
Audiobook reviews struggle with oversimplified rating and overwhelming text fields.
Audiobook reviews struggle with oversimplified rating and overwhelming text fields.
Audiobook listeners want to share their thoughts and feelings about books. But most platforms offer a basic 5-star rating system and a blank field. The rating simplifies the review process but doesn't provide valuable insight. A bank field requires a higher effort to fill out, which makes listeners feel overwhelmed, resulted a higher chance to drop the review process.
Audiobook listeners want to share their thoughts and feelings about books. But most platforms offer a basic 5-star rating system and a blank field. The rating simplifies the review process but doesn't provide valuable insight. A bank field requires a higher effort to fill out, which makes listeners feel overwhelmed, resulted a higher chance to drop the review process.
Role & responsibility
Design and iterate the branding, content, components, and user experience.
Work as the founding designer, collaborate closely with the CEO, CTO, and PM remotely.
Plan, prototype, and conduct various user testing and research to shape the product and assist the decision-making.
Impact
Impact
✅
✅
Validated the effectiveness of the review flow, with 100% of users reporting ease of navigation and a 90% completion rate.
Validated the effectiveness of the review flow, with 100% of users reporting ease of navigation and a 90% completion rate.
👓
👓
Conducted research and analysis to uncover insights overlooked by competing products, shaping the unique product mission & vision.
Conducted research and analysis to uncover insights overlooked by competing products, shaping the unique product mission & vision.
🎨
🎨
Designed the audiobook review that collects critical user data for future feature scaling.
Designed the audiobook review that collects critical user data for future feature scaling.
Role & responsibility
Design and iterate the branding, content, components, and user experience.
Work as the founding designer, collaborate closely with the CEO, CTO, PM, and the developer remotely.
Plan, prototype, and conduct various user testing and research to shape the product and assist the decision-making.
Conduct user journey flow, storyboard, and other UX artifacts to understand users and communicate with stakeholders.
CHALLENGE
CHALLENGE
Serve our users, not every user
Serve our users, not every user
I conducted interviews and tests with stakeholders at times. Sometimes, stakeholders wanted to apply all user insights without thorough analysis. To solve this, I needed to make sure the tests are well recorded and take detailed notes at all times. This was crucial for reminding stakeholders about our previous research and testing work. And I could show the data as many times as needed. Additionally, I iterated on the testing methods and interview questions to gather different kinds of data for further validation. This approach allowed me to back up my points with solid data and help stakeholders better identify major patterns that aligned with our values and goals.
I conducted interviews and tests with stakeholders at times. Sometimes, stakeholders wanted to apply all user insights without thorough analysis. To solve this, I needed to make sure the tests are well recorded and take detailed notes at all times. This was crucial for reminding stakeholders about our previous research and testing work. And I could show the data as many times as needed. Additionally, I iterated on the testing methods and interview questions to gather different kinds of data for further validation. This approach allowed me to back up my points with solid data and help stakeholders better identify major patterns that aligned with our values and goals.
TARGET USERS
TARGET USERS
Romance audiobook listeners
Romance audiobook listeners
“Listen to romance audiobook is an escape from the world for me.”
👀 Constantly looking for the next book to listen to.
👀 Constantly looking for the next book
to listen to.
📚 Active in a romance book community.
📚 Active in a romance book community.
⭐️ Narration performance is a big factor in how she feels about the book.
🎤 Narration performance is a big factor
in how she feels about the book.
🥺 Wants to share about her feelings with the world when finished a book.
💜 Wants to share about her feelings
with the world when finished a book.
😥 Doesn't know how to articulate thoughts in a book review that only has a big blank field.
😥 Doesn't know how to articulate thoughts in a book review that only has a big
blank field.
😥 Doesn't know how to articulate
thoughts in a book review that only
has a big blank field.
DISCOVER
DISCOVER
Insights from previous user interviews from romance book readers
Insights from previous user interviews from romance book readers
We started the product development by focusing on the romance book readers to better tailor the review to this genre. My work started to familiar myself with the user insights we already had from previous interviews. There were no recordings and the notes were not detailed. But there were these repeated insights from 17 interviewrs:
We started the product development by focusing on the romance book readers to better tailor the review to this genre. My work started to familiar myself with the user insights we already had from previous interviews. There were no recordings and the notes were not detailed. But there were these repeated insights from 17 interviewrs:
🔎
🔎
Users want to be able to search or avoid books by story elements, narration style, tropes, and mood.
Users want to be able to search or avoid books by story elements, narration style, tropes, and mood.
👎
Users don’t like to leave feedbacks on a blank field. And the rating on Audible feels limited.
Users don’t like to leave feedbacks on a blank field. And the rating on Audible feels limited.
🚫
Users don't want book recommendations by popularity or promotion.
Users don't want book recommendations by popularity or promotion.
None of the platforms have detailed narration review
None of the platforms have detailed narration review
To see how other audiobook/book platforms do with the review feature, I did a competitive analysis:
To see how other audiobook/book platforms do with the review feature, I did a competitive analysis:
Audible
😐
Basic 5-star rating for book and narration performance.
😐
A big blank field.
😐
Recommendations are based on promoted or popular trends.
Storygraph
✅
Detailed book review, very close to what we want.
✅
Spoiler check.
😐
No narration detail.
😐
All questions on one page. Lots of text, may make users feel overwhelming.
😐
Navigation is not intuitive at times.
Goodreads
😐
Basic 5-star rating for book but no narration.
✅
Spoiler check
😐
No narration detail.
😐
A big blank field
😐
Outdated UI
Audible
😐
Basic 5-star rating for book and narration performance.
😐
A big blank field.
😐
Recommendations are based on promoted or popular trends.
Storygraph
✅
Detailed book review, very close to what we want.
✅
Spoiler check.
😐
No narration detail.
😐
All questions on one page. Lots of text, may make users feel overwhelming.
😐
Navigation is not intuitive at times.
Goodreads
😐
Basic 5-star rating for book but no narration.
✅
Spoiler check
😐
No narration detail.
😐
A big blank field
😐
Outdated UI
Romance.io
😐
Basic 5-star rating for book but no narration.
✅
Spoiler check
✅
Genre-specific rating: Spicy rating
✅
Has very detailed trope info.
😐
No narration detail.
😐
Outdated UI
Through the analysis, I noticed none of the competitors asked about narration production and narrator details in the review. However, through the interview insights, we know that audiobook listeners care about narration a lot. By focusing on the narration details more, we will have a better experience for the users.
Through the analysis, I noticed none of the competitors asked about narration production and narrator details in the review. However, through the interview insights, we know that audiobook listeners care about narration a lot. By focusing on the narration details more, we will have a better experience for the users.
DEFINE
DEFINE
What we want from users in the review
What we want from users in the review
Before starting to design a review prototype to further conduct user testing, we finalized what we wanted to know from the users in the review, based on the previous interviews and the competitive analysis.
Before starting to design a review prototype to further conduct user testing, we finalized what we wanted to know from the users in the review, based on the previous interviews and the competitive analysis.
1️⃣
1️⃣
Feelings they have from the book.
Feelings they have from the book.
2️⃣
2️⃣
More detailed narration review and rating.
More detailed narration review and rating.
3️⃣
3️⃣
How much did the narration and book match their preference?
How much did the narration and book match their preference?
4️⃣
4️⃣
What influenced their rating?
What influenced their rating?
5️⃣
5️⃣
What do they wish to be able to search to find this book?
What do they wish to be able to search to find this book?
6️⃣
6️⃣
Collect Trigger Warnings?
Collect Trigger Warnings?
3️⃣
In-depth book ratings, what made the book great or not so great.
5️⃣
Collect Trigger Warnings.
6️⃣
Specific Genre Ratings, spicy level, funny level…
DEVELOPE
DEVELOPE
What words will be able to cover the moods of books
What words will be able to cover the moods of books
The review needs to collect what feelings users have from the book. To figure out what words options we should provide, I went on Reddit’s book communities, and collected how readers describe the feelings they had from books. After showing to the team, we grouped the similar words together and came up with various words to describe feelings as a start to test with.
The review needs to collect what feelings users have from the book. To figure out what words options we should provide, I went on Reddit’s book communities, and collected how readers describe the feelings they had from books. After showing to the team, we grouped the similar words together and came up with various words to describe feelings as a start to test with.
Experimenting with various presentations of the questions and asking for feedbacks
Experimenting with various presentations of the questions and asking for feedbacks
Now the mood words options are set, I started to ideate how to present the review questions and how would users interact with it.
Now the mood words options are set, I started to ideate how to present the review questions and how would users interact with it.
During the process, I asked for feedbacks from the stakeholders for a couple of times, which helped me a lot with understanding more of their perspectives. Then prototype was finalized and ready to be tested.
During the process, I asked for feedbacks from the stakeholders for a couple of times, which helped me a lot with understanding more of their perspectives. Then prototype was finalized and ready to be tested.
1 question per card. Help users focus on one question at the time.
Progress bar to let users know how many questions to expect.
"x" indicates users are free to exit anytime.
Consistent card size for the ease of navigation
Hide the blank field. If users want to share more, they can click “Tell readers more”.
Hide the blank field. If users want to share more, they can click “Tell readers more”.
Progress bar to let users know how many questions to expect.
"x" indicates users are free to exit anytime.
Consistent card size for the ease of navigation.
Hide the blank field. If users want to share more, they can click “Tell readers more”.
I also divided the review into 2 parts to make users feel less overwhelmed. This is to ensure users will have a higher chance to complete the first part of the review. The first part contains reviews of the book and narration performance. The second part is to collect book tags and other insights for future feature.
1 question per card. Help users focus on one question at the time.
I divided the review into 2 parts. The first part contains reviews of the narration and the book. The second part is to collect book tags and other insights for the future feature. By doing this, I wanted to see if it makes users feel less overwhelmed, and see if it will lead to a higher completion rate for the first part.
I divided the review into 2 parts. The first part contains reviews of the narration and the book. The second part is to collect book tags and other insights for the future feature. By doing this, I wanted to see if it makes users feel less overwhelmed, and see if it will lead to a higher completion rate for the first part.
DELIVER
DELIVER
Define what I want to know from the testing
Define what I want to know from the testing
I conducted moderated testing to ask users to have a book in mind and run through the review. The goal is to know:
I conducted moderated testing to ask users to have a book in mind and run through the review. The goal is to know:
🤔
🤔
How will users interpret questions and answers?
How will users interpret questions and answers?
🤔
🤔
Can users navigate and interact with the review as I expected?
Can users navigate and interact with the review as I expected?
🤔
🤔
How do they feel about the review process? Is it too overwhelming for them?
How do they feel about the review process? Is it too overwhelming for them?
🤔
🤔
Is there anything else they would like to express about the audiobook?
Is there anything else they would like to express about the audiobook?
1️⃣
Users found some words options are difficult to understand. And still need to add more words options.
2️⃣
Users didn't know if they feel neutral about the elements they could skip.
3️⃣
Users only focused on trigger warnings and ignored plot points and ingredients
4️⃣
No users clicked “Tell readers more” to fill out more feedbacks.
1️⃣
Based on the words users added in the testing, I reworked on the word options and simplified the words.
2️⃣
Combined the elements to the overall book rating. The second part will only show when users rate very low (below 2) or very high (5).
3️⃣
Separate the book element tags from the trigger point card.
4️⃣
A dedicated card for users to fill out. To avoid overwhelming the users, this part is optional.
Insights that improved more than half of the review design
Insights that improved more than half of the review design
Each testing was 30 mins. Users were asked to have a book they recently read in might and go through the review by themselves. Then I asked them to do it again, with questions about each review card, to get their feedbacks.
After 5 tests with romance book readers, there were repeated feedbacks. I iterated the design and ready for it to be tested again.
Each testing was 30 mins. Users were asked to have a book they recently read in might and go through the review by themselves. Then I asked them to do it again, with questions about each review card, to get their feedbacks.
After 5 tests with romance book readers, there were repeated feedbacks. I iterated the design and ready for it to be tested again.
1️⃣
Users found some words options are difficult to understand. And still need to add more words options.
2️⃣
Users didn't know if they feel neutral about the elements they could skip.
3️⃣
Users only focused on trigger warnings and ignored plot points and ingredients
4️⃣
No users clicked “Tell readers more” to fill out more feedbacks.
1️⃣
Based on the words users added in the testing, I reworked on the word options and simplified the words.
2️⃣
Combined the elements to the overall book rating. The second part will only show when users rate very low (below 2) or very high (5).
3️⃣
Separate the book element tags from the trigger point card.
4️⃣
A dedicated card for users to fill out. To avoid overwhelming the users, this part is optional.
1️⃣
Insight
3 out of 5 users felt difficult to understand some word options. Users suggested other words that might be missing.
Iteration
Simplified words and added more words options. All users were satisfied with the word options in the next 5 testings.
2️⃣
Insight
How all users clicked all thumbs up shows even when they don't like the book showed that this design will not provide valuable data.
Iteration
Combined the elements card with the book rating. The second part only shows when users rate above 4 or under 2 to help users express what worked well or not well for them.
3️⃣
Insight
All users focused on helping other users avoid books. No users mentioned things that will help others to find the book.
Iteration
Let the original questions focus on trigger points and create another question for adding tags for the book.
4️⃣
Insight
No users clicked “Tell readers more”. But we do want users to express more through their own words.
Iteration
Removed all "Tell readers more". Created a dedicated card for users to fill out. To avoid overwhelming the users, this part is optional.
Improve the testing experience and see what users will do when reviewing books they dislike
Improve the testing experience and see what users will do when reviewing books they dislike
To improve the testing experience and discover more of what users might do, I asked for what book they wanted to review in advanced. Then I provided different prototypes based on one, two narrators or more than 2 narrators to improve the testing experience. Users were also asked to review books they don't enjoy.
After 5 testing, some review questions were validated and ready to handoff. But there are still some user pain points that were discovered:
To improve the testing experience and discover more of what users might do, I asked for what book they wanted to review in advanced. Then I provided different prototypes based on one, two narrators or more than 2 narrators to improve the testing experience. Users were also asked to review books they don't enjoy.
After 5 testing, some review questions were validated and ready to handoff. But there are still some user pain points that were discovered:
1️⃣
Most users would rather not continue the second part of the review because they didn't know how many more questions there are.
2️⃣
Users expressed that they want to know how much the trigger warnings were mentioned in the book.
3️⃣
Users wonder if there is no intimate content in the book at all, can they rate 0?
4️⃣
Users mentioned other elements they would want to search to find books, but when seeing this, they only thought of tropes.
1️⃣
Insight
More than half of the users express this is where they might drop the review because they didn't know how many more questions there are.
2️⃣
Insight
Half of the users wanted to know how much the trigger warnings were mentioned in the book when searching for the book to read/listen to.
3️⃣
Insight
1 User wondered if there is no intimate content in the book at all, can they rate 0?
Quantitative data that finalized the review and convinced stakeholders not to make “every” user happy
Quantitative data that finalized the review and convinced stakeholders not to make “every” user happy
At this point, it was clear that the questions we still have will not be solved efficiently by running individual testing or interviews. , I worked with the PM to draft a survey for romance book lovers. We wanted to know:
At this point, it was clear that the questions we still have will not be solved efficiently by running individual testing or interviews. , I worked with the PM to draft a survey for romance book lovers. We wanted to know:
🧐
🧐
What users wish to be able to use to search by for a book?
What users wish to be able to use to search by for a book?
🧐
🧐
What are the things most users care about when it comes to narration of the audiobook?
What are the things most users care about when it comes to narration of the audiobook?
2 users expressed that they would like to know the intimate details or certain plots in advanced in the book. It's only 2 testers out of 10 testings I have done. One of the stakeholders wanted the review to be iterated for this insight. To better validate this before the iteration, this question was also added into the survey:
2 users expressed that they would like to know the intimate details or certain plots in advanced in the book. It's only 2 testers out of 10 testings I have done. One of the stakeholders wanted the review to be iterated for this insight. To better validate this before the iteration, this question was also added into the survey:
🧐
🧐
If a book contains explicit content, how much users want to know in advance?
If a book contains explicit content, how much users want to know in advance?
The survey reached to 30 people in the romance audiobook community. And I learned that:
The survey reached to 30 people in the romance audiobook community. And I learned that:
⭐️
⭐️
50% of users want to search by tropes. The rest of the search wishes were random.
50% of users want to search by tropes. The rest of the search wishes were random.
⭐️
⭐️
70% of users care about the narration if it's emotional, authentic, fit to the character and with proper pronunciation.
70% of users care about the narration if it's emotional, authentic, fit to the character and with proper pronunciation.
⭐️
⭐️
62% of users don't want to know any intimate or certain plot content before listening to the book.
62% of users don't want to know any intimate or certain plot content before listening to the book.
And the survey result was very helpful. It helped the team to align with how to finalize it. I iterated the design according to the insights. It felt close to the goal. I iterated the design and did 5 more moderated usability testing to validate it before handoff.
And the survey result was very helpful. It helped the team to align with how to finalize it. I iterated the design according to the insights. It felt close to the goal. I iterated the design and did 5 more moderated usability testing to validate it before handoff.
RESULT
RESULT
Detailed narration rating
Detailed narration rating
Users can rate by the narrator and share how they think about the
performance, voice quality and production details.
Users can rate by the narrator and
share how they think about the
performance, voice quality and production details.
If more than 2 narrators, it would be a general cast rating to avoid too many questions that might overwhelm the users.
If more than 2 narrators, it would
be a general cast rating to avoid too many questions that might
overwhelm the users.
Users can rate by the narrator and share how they think about the performance, voice quality and production details.
If more than 2 narrators, it would be a general cast rating to avoid too many questions that might overwhelm the users.
Rating with a legend
Rating with a legend
Specify what each rate means to make sure users have the same understanding of the rating standard.
Specify what each rate means to make sure users have the same understanding of the rating standard.
What made the book stand out or ruined the book
What made the book stand out or ruined the book
Easy rating but knowing why users feel the book is great or bad.
The second part of the question will only show when users rate below 2 hearts or 5 hearts.
Easy rating but knowing why users feel the book is great or bad.
The second part of the question will only show when users rate below 2 hearts or 5 hearts.
Optional blank field with spoiler check
Optional blank field with spoiler check
Optional blank field, encourages users to share more, but also enables them to skip.
Spoiler check to prevent other users from getting spoiled when going through reviews.
Optional blank field, encourages users to share more, but also enables them to skip.
Spoiler check to prevent other users from getting spoiled when going through reviews.
Let users know what to expect next
Let users know what to expect next
Separate the review questions into 2 parts. Improving the completion rate by putting important questions into the first part.
Specify how many questions for the next part and what is the purpose for the users to go through the questions, so users will know what to expect.
Specify how many questions for the
next part and what is the purpose for the users to go through the questions, so users will know what to expect.
Separate the review questions into 2 parts. Improving the completion rate by putting important questions into the first part.
Specify how many questions for the next part and what is the purpose for the users to go through the questions, so users will know what to expect.
Detailed triger warning tag
Detailed triger warning tag
Users can mark if the trigger warning is off, on page or main plot.
This will help other users make better decisions when looking for the next book.
This will help other users make better decisions when looking for the next book.
Users can mark if the trigger warning is off, on page or main plot.
This will help other users make better decisions when looking for the next book.
💙
💙
100% users find it easy to navigate in testing.
100% users find it easy to navigate in testing.
✅
✅
✅
90% completion rate for the book review.
90% completion rate for the book review.
🙌🏼
🙌🏼
🙌🏼
80% of participants from user research were excited and signed up for the launching notification.
80% of participants from user research were excited and signed up for the launching notification.
Next steps
Next steps
With the expectation to release the MVP this year, here are the priority next step for me to help Booklet achieve its goal:
With the expectation to release the MVP this year, here are the priority next step for me to help Booklet achieve its goal:
🎯 Create MVP user flow.
🎨 Design the home page and book profile page.
🧪 Continue testing and iteration.
🎯 Create MVP user flow.
🎨 Design the home page and book
profile page.
🧪 Continue testing and iteration.
TAKEAWAYS
TAKEAWAYS
Using research and metrics to aid communication
Using research and metrics to aid communication
I am very lucky to work with a team that is very supportive and truly care about what we do. However, sometimes it's difficult to align everyone. Not everyone is familiar with how to approach with a user experience aspect or how to balance UX with business goal, vision, and personal insight. I learned that I should speak out more but speak with structure and backed up by a research, metrics, or technical reasons. I wish I have done it earlier in the project.
I am very lucky to work with a team that is very supportive and truly care about what we do. However, sometimes it's difficult to align everyone. Not everyone is familiar with how to approach with a user experience aspect or how to balance UX with business goal, vision, and personal insight. I learned that I should speak out more but speak with structure and backed up by a research, metrics, or technical reasons. I wish I have done it earlier in the project.
This is why I love what I do
This is why I love what I do
When doing research and testings for Booklet, I experienced how users become excited and passionate about what we are doing when it’s solving a problem they've experienced. A lot of them even refuse the incentives and want to keep testing with us. I learn that this is what a good product should do, and to build a product like this and create a product like this, is the purpose of what I do, and I love it!
When doing research and testings for Booklet, I experienced how users become excited and passionate about what we are doing when it’s solving a problem they've experienced. A lot of them even refuse the incentives and want to keep testing with us. I learn that this is what a good product should do, and to build a product like this and create a product like this, is the purpose of what I do, and I love it!