product design case study

REBUILDING AND LOCALIZING THE NEW KOBO E-READER MOBILE APPS

2019-2020

12 min read

Kobo-hero-image@2x

10 years ago, e-book readers looked out to be a thing of the future. When we got it, they seemed clunky and impersonal, but with time, little by little, they got better. 10 years forward, we have them in our phones, stringing along with us while commuting, before going to sleep or even during short coffee breaks.

Here is my experience on the next story, where we take a book reader made for all and made it for a few, so that those few could enjoy it reading even more. It’s how we localize the Kobo e-book reader app –made for readers around the globe– to the readers in Japan.

Note: to comply with my non-disclosure agreement, I have blurred and obmitted confidential information to comply with my non-disclosure. The information in this case study reflects my personal view and experience and does not necessarily reflect Rakuten Kobo’s viewpoint.

Background

Kobo is an e-book leading company that was founded in Canada in the late 2000s, expanded to the Americas, Europe and is now available in most of Asian Markets as well. The company was bought by the Japanese e-commerce giant Rakuten and is now a core part of its ecosystem.

The initial business goal was to integrate the product in the Japanese market, as well as Rakuten values and their core technology. The next step as a Designer was to completely understand the market, customers and their needs as users.

My role

I led a team of UX & UI Designers for this project, whose focus was to understand the local market and audience, rebuild and redesign the mobile & tablet apps. I worked alongside the Product Manager, Product Owner and 4 Developers.

The project execution started late 2018 and finished early 2020 for iOS.
The project is yet to be launched for Android.

some of my tasks were

1. Gathering early Market & Customers Insights
2. Ideation & Leading Workshops
3. Planning and Scope Definition
4. Team Leading, Management and Coordination within different departments and overseas design teams
5. Create the User Experience & First Interfaces

1. THE CHALLENGE

Localizing and redesigning an app from a global solution

Localizing a service presents very different challenges than creating a new service or preparing it to scale.

The Kobo app already had a solution that was working for “everybody” –a global standard user– but it wasn't working as well for the new Japanese market. Despite that, it became the number 1 selling market in the platform in a matter of months. Even so, for these new users, the product wasn't fulfilling their needs & expectations. They were accustomed to a different experience: they read different content than overseas users and the target demographic was extremely different as well. At this point, Kobo Japan already knew that they needed to localize the product and there were already too many hints on what we needed to do in order to maximize the users satisfaction and enhance the business local growth.

The general audience

In a country that creates so much literature and manga content and with more than 70 competitors in the market, we needed to differentiate ourselves by changing the experience and service. In comparison to the rest of Kobo users, the Japanese users in average:

  • Read more manga and comic content than written books
  • Spent commuting a lot of hours in a weekly basis
  • Younger audience (20-mid 30s)
  • Had a different written language, culture and in summary, way to read and understand reading

High level goals

  1. Create a library to encourage users to read more comic content
  2. Enhance the comic reading experience
  3. Create an app that the users feel it's their own and it's an extension of themselves

I also had a personal challenge: to understand a new culture.

2. What I knew from the context

Understanding the culture and market

When I joined the company, I needed to quickly onboard on a new culture and market. Not knowing the language was just the first of the hurdles. Understanding the Culture was going to be the most challenging one. I didn't have a lot of time to onboard, so I focused on figuring out what were the specific points in which the users read, use their books differently and how their habits differ.

Habits diagram

I summarized from different internet sources and with the great help of the Product Manager and the UX Designer in my team. The next are the key differences that gave shape to the new platform.

Understanding the culture

Social differences
  1. Hard working culture that values less leisure time and spend less time with their family
  2. Strong visual history and manga or comics are deeply ingrained in their culture for centuries
Ukyo-e painting

Traditional wood paintings or Ukiyo-e have represented daily lives from the 16th up until the end of the 19th century and are one of the main reasons why visual tradition has been well conserved and was a big part of the culture that ended up inspiring other tipes of visual communication including Shin-hanga (or new prints), manga (or comics) among others. Kanegafuchi no Yurai (History of Kanegafuchi), c. 1848 From the play Kaidan Sumidagawa.

       3. Longer commute in highly dense areas and public transportation

Average time commuting in Japan

Written culture differences

  1. A more complex written system: Japanese scripture is a logographic written system that comprehends 3 different writing systems: Kanji (Chinese adopted system with more than 4000 characters), Hiragana (autochthonous syllabic system) and Katakana (a syllabic system created to represent foreign phoneme).
Japanese script sample describing 'Hanami'

The word ‘hanami’ (fireworks) in the 3 different script systems. In order from left to right, Kanji, Hiragana and Katakana.

To check more information about the Japanese writing systems, check Shoko Mugikura’s beautiful article Japanese, a beautifully complex writing system.

  1. Reading direction goes from top to bottom and then right to left. Since internet exposition, left to right (as English) has been popularized a lot
  2. The same books tend to be shorter in Japanese than English
  3. Characters do not have multiple weights (they tend to have 2: regular and bold) and images, colors and change in size are commonly used to represent different visual and typographic hierarchy
Street signs at Night in Shibuya

Nightlife in Shibuya, Tokyo. The streets are busy where signs compete ferociously for the walkers’ attention.

Understanding the market

  1. Integration in Rakuten. Rakuten is the largest e-commerce and ecosystem service company in Japan. It currently has a user base larger than 100 million users according to statista. Marketing was running multiple strategies to attract these users and Kobo needed to adapt its philosophy, brand image, technology and operations
  2. Some user behavior e-commerce trends for East Asian markets includes higher interaction with banners, higher Average Time on Site and users tend to look more at the content than skim through it as the Western counterpart

After I had a better idea of the culture and market, I created the roadmap and laid out the blueprint of how we would tackle the project.

4. roadmap & design process

From large problems to solutions

Now that we understood where the big problem relies, I worked with the product and project management teams to update the roadmap and prioritize user research where we needed. This was going to be a more than a year project –rebuilding the whole app took its time– so it was vital to reduce complexity and invest in specific pain points and substantial opportunities of improvement.

After 2 prioritization workshops and further work from the product management side this was the final roadmap:

Project Schedule

Project Gantt Chart (emphasis in Library)

In this use case, I’ll address the improvements and processes in while redesigning the library.

Initial Design Process

Product design process

This is an evolution from the 5D Product Design process refined by Altuit that I adjusted to the needs of the project.

5. kickoff

Prioritizing User Research

User research can take a lot of time and resources. It was key that we identified where our needs were in order to prioritize correctly.

Right after understanding the market and cultural differences better, I needed to verify the assumptions that we had collected to understand the current behavior of the product, pain points, goals and expectations from the users.

We divided the research into 3 different blocks:

  1. Study environment habits in the beginning and after the app release
  2. Prioritize app usage outside of reading user research (app sections)
  3. Increasingly improve while reading habits experience (revising and creating new features)

6. user research

Understanding user behavior, motivation and pain points

What can we offer to the users in the app other than reading?

1. Summarizing Data

Findings from usage data and major demographics to the Product and Business teams in Kobo Japan.

That is a more profound redefinition of what I mentioned previously as app usage outside of reading. To understand more, me and the UX Designer in the team summarized quantitative data and crossed it with qualitative data via interviews and finished with a user activity.

  • About 50% of the content are manga books compared to less than 10% overseas

  • A small amount of users have thousands of books and they made up for a large part of the purchases

  • From the Collections feature (users could create one personal collection and add books to it), people were creating collections just for their names. Besides from that, they were barely being created

  • Search was mostly used to find books directly, rather than searching by authors, genres or other related content

  • More than 90% of the users were using the first tab of the library: Books. Others, such as author views were barely used in Japan: we assumed that it was more suited for Novel and Business books rather than comics

  • Continue reading and Recently added were the most used lists in the app

  • Libraries that had a lot of manga books were really heavy and could slow the app performance

  • Other library usage data related to tab structure access and filtering

Library structure

Current library structure

2. User Interviews

We interviewed 7 users to understand which issues would they find in their book libraries and if there were any trends or insights that could help us to improve the current experience. These were some of their findings:

  1. Users had a hard time to find their favorite genres. It was easy to find the books in the store than in their library

  2. Users identify books by their covers rather than by their names. This could be due to a culture that has a higher visual recognition or because the titles can be written with different writing systems ("or alphabets") which makes finding these titles more difficult and less recognizable

  3. Authors are not as relevant to recognize a title, especially considering as representative for manga books

  4. Users didn't find easy to continue reading different books in the same series

3. The Activity

In order to gain a deeper understanding of how the users search and organize their content, we laid out a card sorting activity. The cards contained different actions and categories of books and content.
We proposed different scenarios and asked the participants how would they locate a book, organize their content to begin reading.

Before starting the activity, we asked the users if they knew how large their library was, the type of content they usually read and if they were using other readers or library managing apps. That would help us understand their mindset behind their decision making.

Card sorting activity: User participation
Card sorting activity: Categorization

Cart sorting activity was done with 7 users where they had to organize first their book library as what they’d do in their bookshelf, how would they improve it and then we’d specify some cases they’d have to portray to understand how they would build their libraries based on them

Some of the findings from the activity were:

  • Users expected to read as soon as purchased and expected to find their books without any steps in between

  • Users wanted to continue reading manga series as soon as the next volume is available and expected to browse the content grouped by series

  • We verified that searching by title, filtering by "on device" or "downloaded", and browsing by author had less impact on how the users would access their content

  • Filtering/browsing by sub-genre seems rarely used, despite it was one of the requests found in the interviews. Only users with large libraries would use filtering often and they would probably need a more advanced or a detailed version of how we had it

  • Fewer predefined categories might help users find their content more quickly

Once the session was over, we had an initial idea of which categories and actions are more useful to improve the Information Architecture of the app.

Filter and sort legacy menu

The current library had quick filters. These didn’t work well for users who had a large library who are the ones using it. We needed to optimize it for different personas.

7. reshaping the vision

Reframing the problem

The larger the library becomes, the more complex problems users inherit. Every user will have a more enjoyable reading experience by simplifying the library for all users and providing specific tools for the expert ones.

8. solving problems

Redefining the service: creating a unique library for each user

To provide unique solutions for the different users that were using our app, I reframed the problems as challenges and created hypothesis to test and implement them. These are some of the major solutions that introduced game-changing improvements.


Problem

It is difficult to find books and manga inside series

opportunity

How Might We help users finding content in their series as fast as possible?

hypothesis

We believe that grouping and optimizing the current library structure will provide a more intuitive way to find and access their content easily.

how we investigated and implemented it

We created 3 different versions that organized the library in a different way:

  1. The first version. Following the users request, we divided the library by the main book categories: novels, manga and magazines
  2. The second version. Illustrated the business models: purchased books, borrowed and uploaded books
  3. The current (control) version. Contains collections, books, authors and series
Library categorization versions diagram

We ran a Usability Testing session with 7 active users in which we found:

  • The content wasn't particularly confusing or hard to get either way
  • The users did not access genres in which they didn’t have any books, meaning that these tabs were useless most of the time
  • The users still missed the capability of organizing their content how they wanted, even if we found out in our data that most of the collections were done to create specific series (for an instance, a lot of users created a collection called "Harry Potter" and then proceeded to add Harry Potter books to the collection).
Control version: current model
Version 1: business model
Version 2: book categories model

Through another round of iteration and feedback, it was clear that we needed the business model version in the top level navigation, but still needed an extra navigation level to grant access to different content types.

Library structure diagram

Giving the bookworms the keys to the library

Up until this point, we had figured out how the users could access and add content, which worked great for most users.

What about those who had thousands of books?

Those that read one book after another, one comic after another, non-stop? These were our most loyal users. After these changes, we still had not solved the largest issue for them: the libraries were too heavy (a comic book is about 20x as large in size as a text book with the same pages) and looking for content in a large library was still painful.

Aside from that, some manga books have with similar covers, and users can forget their names and some titles have a lot of characters, and were not as easy to search for.

Problem

Users with large libraries can’t find their content

opportunity

How Might We help users with large libraries finding their content?

hypothesis

Creating an advanced filter will give users with a large library a better way to refine and find their content.

how we invesigated and implemented it

We wanted to facilitate the work for the heavy users and not hurt the middle and novice user. By creating and advanced filter, we were giving more control to the heavy users but we didn't know how the others one would react, considering that we didn't have usage data, just opinions from our previous interviews and customer care.

The current control version had all the filters in one sheet in which the user could select only 1 option at a time. I ideated an advanced filter version that split them into 2 sections or steps and gave more control for advanced users.

Legacy app: old filter

Old app filter (sort & filter in a floating menu)

Advanced filter, 1st level wireframe

New filter: 1st level, summary

Advanced filter, 2nd level wireframe

New filter: 2nd level, filter detail

Advanced filter wireframes for tablet

Tablet version had the 2 steps in 1 modal dialog

After the testing session and interviewing the users we found:

  1. Users with small and mid size libraries (up to 200 books) almost never used the filters, they found content by scanning/browsing or searching
  2. Users with small and mid size libraries used sorting with relative frequency (mostly to switch from recently added to a character ordering)
  3. Users with large library wanted to know if their purchased content was already downloaded or not. It was specially useful for re-reads
  4. Users with large libraries found that having a more granular control of their libraries helped them to find their content quickly
  5. Some users reported that it was difficult to know then the filters were applied after reloading or re-entering into their library. I fixed it by improving visibility and recovery options

In the end, the 2 steps filter was successful for large library users and didn’t take anything away from the rest of users. Likewise, having a separate sort option helped the rest of the users focusing on sort options and omitting seeing filtering options that were not relevant for them.

Problem

Users can only find book results (and not collections, series or authors) while searching

opportunity

How Might We show other suggestions or results aside from books while searching?

hypothesis

Dividing search suggestions by categories will help users to identify their content faster and provide multiple category results when searching.

how we invesigated and implemented it

The premise was simple: adding different categories to the search prediction and on the page results as well. The initial order came from top search usage, starting by history searched term (users may book the same book more than once), series, books, authors, collections and uploaded books. Also as a result of separating the Library search experience from the Store search experience, we were able to target the search better and provide better contextualized results.

Search wireframes for the library

Search contextualized only for ‘My library’ section

Search wireframes: first time

Once searching, first time experience or history suggestions

Search prediction wireframes

Search suggestions are grouped in small sets to be easily reached

Search wireframes: results

For search results, a similar categorized structure showing all results and full list view

After the usability test sessions we found out that:

  • The time to find books was up 0.5 seconds
  • The time to find TV series was faster than the previous version that could only find books
  • Users that were looking for books in a series would often tap in the series to see their books and then find the latest content
  • Some users would search predominantly different categories. There is an opportunity to improve suggestions by predicting a better order in categorization
  • Overall positive feedback review
9. other changes

Polishing until the last book

These were the initial 2 big changes and many more came. After all, changing the structure of your library can mean that the shelves can break or the books fall flat. I needed to make sure that the changes made a big impact.

Refined sort, filter and view types access

The main structure of the library was working well now, but in the new library, the visualization and grouping options didn’t match 1-to-1 as they did previously. Here’s a quick view mapping of these options:

Sort, grouping and visualization types diagram

Sort, grouping and visualization options

To address them in the most accessible, yet engaging way, I iterated and tested internally with users in the company. These were the final views:

First version: sort and filter options

Sort and filters (filters would have grouping inside)

Filter-tab-version-2

Sort, grouping and filters

Final filter, sort and view options bar version

Filters, sort and visualization options. This was the final version

More books per screen

Changed the default book grid from 2 columns to 3 columns on smartphone to show 150% of the original content:

Old library mockups

Old (screenshot)

New wireframes for the library home

New (more)

New series view

Added a series view list and detailed content

Author view wirefames

Series view

Second level library wireframes (inside author view)

Inside series

Structure

Information architecture & taxonomy

Aiming to improve recognition while keeping the library as simple as possible, we made some small improvements, aimed to have visibility of the status of the book progress, ownership and download state

Ownership states per book diagram

This is the final representation of the states over the book main information.

Book components Information Architecture

This is the final structure that shows the minimal information that a book needs to be recognised, download, read states, progress and actions access.

10. design system

Creating a Scalable Design system for multiple services

Phases

We started working in the local design first, and half way through, we started working with the other 2 reader services and design teams overseas. We synchronized and started working in different features that the others would optimize and implement and vice-versa.

The creation process

We put a lot of effort and emphasis towards standardizing UX and UI creation processes. We had a Designer in charge of creating the main components in a centralized library using Figma. Each one of the teams used these components to customize them and apply their styles to fit each service brand and needs.

In the end, we created components, mockups and user journeys by customizing the Shared Basic Library. We also discussed and added components back to that library for the other teams to use.

This is a quick mock of how we build our libraries:

Design libraries distribution and dependencies

Examples of the library customization:

Empty library. Unbranded (white label)
Empty library. Branded

Customization of the Top bar and Bottom navigation bar from the Shared Basic to the Kobo Japan Library.

10. design system

Final mockup samples

Note: The following mockups are in progress and the shown samples do not reflect the final UI version

Group & views

Top page in the library for smartphone
Group and view option list

My Books - Tablet view

Library for tablet

Library search

Book search video
Recommendation type list

Book Details

Book detail page
Detail page for books: tablet landscape
11. wrapping up

Conclusion

Challenges

The biggest initial challenge was to be able to dive in deep into another culture that represented so many changes on how the users behave and what would be the best User Experience for them.

It was also challenging to lead multiple Designers working with different features, sometimes by themselves, sometimes in squads while designing at the same time.

Another big challenge was coordinating 3 different products that were sharing design and development core despite having different user needs. This represented a lot of going back and forth despite time zone differences. Creating a common scalable design system was challenging to say the least.

Takeaways

Working across geos

Some of my major learnings was how to set different UX approaches while leading a team into one direction. It took a multiple team effort, lots of trials and errors, redefining and refining processes.

Creating a multi-team scalable design system

Creating a multi-scalable design system was very interesting and required major coordination between stakeholders overseas and also implementation inside the team. It also forced me to plan ahead of schedule for possible changes and I had to be able to mitigate some of the issues that it entangled. In the end, it helped us create a more consistent and escalable project that would serve not only us, but other teams and companies.

Dealing with multiple projects

At the time I was also directing other projects from different companies, which made time management critical and made me plan and make accountable my time investment and resource management from the different Designers in my team.

Managing a team and designing

This was my main personal work in progress for most part of the project. I was both a Designer for the project and a Team Lead and Manager of a group of 5 Designers –for this and other projects– working for different companies. It was extremely difficult at times to manage between the 2 different worlds of designing and managing. There is a different mindset that these positions require and, again, time management is a key factor.

After a while, I got a better handle on it but the periods of time or allocation between the tasks that I would require to do fluctuated and it was tiring at times, but definitely a learning experience and an enriching one in terms of not just being able to do one thing but forcing myself to watching issues from different points of view. Being able to talk and emphatize with different stakeholders from different worlds and cultures was an experience that I will treasure.

Next steps for the project

There is still a lot of work to be done. Currently, the iOS app is about to be released and the Android will come in the next few months. Some of the next improvements might be:

 

  • Check different UI approaches with a larger user base (starting with filtering content)
  • Refinement of search prediction through AB testing
  • Refinements on the library through AB testing
  • Crafting and implementing an Animation Guideline to improve the design system (in progress)
  • Multiple UI improvements

Copyright © Jaime Zamorano. 2020
Legal details

Back to top Arrow