Goodreads
Personalizing Goodreads by adding a librarian recommendations feature
Goodreads, the world’s largest online reader's resource, has a seriously flawed recommendation system that requires extensive up-front work from the user to receive suggested books. By adding a feature that connects users with their local librarians to get hand-picked recommendations, I designed a simple and engaging way for readers to find new books and connect with their community.
The Context
Goodreads, the world’s largest online reader's resource, allows users to track the books they’re reading, set reading goals, read reviews from other users, and get recommendations based on their reading habits. Having used Goodreads myself for over three years, I was able to identify a number of ways for the site to be more accessible and user-friendly. I decided to take a closer look at some of the issues I identified and probe further into why they occurred.
The Problem
Many sections of the site are difficult to navigate due to a deficit of simple and effective criteria that users can access to filter through Goodreads’ massive database. Additionally, the site’s recommendation feature requires extensive up-front work from the user to receive suggested books. On top of that, I found that often the books I was recommended weren’t ones that I was interested in reading.
The Objective
I decided to conduct interviews with Goodreads users to understand the issues they were having with the site and its features. I wanted to look closely at how users interacted with Goodreads’ UI and how it may be playing a part in issues of accessibility and efficiency.
The Solution
1. Empathize
Evaluating other book tracking platforms using competitor analysis
Using competitor analysis, I analyzed four book tracking platforms to assess their strengths and weaknesses. Since Goodreads is by far the most widely used book tracker, I chose a variety of tools, big and small, to get a broad sense of the different options on the market. Through my research, I found that customizable tags for searching and cataloguing books were highly successful among users of the competition. Additionally, while some other platforms had different methods of recommending books to users, there were nonetheless still unnecessary hurdles for users to surpass to receive recommendations. Even with successful and engaging UI, these platforms still lacked a simple way to get recommendations to their users.
Interviewing Goodreads users to learn how they use the site
With a clearer sense of the competition, I wanted to speak with users of Goodreads to learn how they felt about the site’s UI and features. I conducted one-on-one interviews with six Goodreads users whose experience with the site ranged from one to twelve years. I thought it was important to speak with users of varying experience to learn if there was an overlap of pain points between new and veteran users, or if issues were more dependent on individual experience.
Whether these readers were new to Goodreads or had used the site for a decade, one feeling was shared among them all: they wanted a better recommendation tool.
Users unanimously agreed that they loved Goodreads’ massive database, and they continued to use the site because, while other platforms had better recommendation tools, cleaner UI, and better search features, none boasted the same selection of books as Goodreads.
Users felt like they couldn’t leave the site that they put so much time into, despite its obvious problems.
Because of this, users found themselves wishing for a recommendation tool that not only offered them personally-tailored recommendations, but allowed them to connect more with other users over their newfound favorites.
Users felt that Goodreads’ recommendations weren’t tailored to their reading habits: the site seemed to pump out a large quantity of inaccurate recommendations instead of fewer, more personalized options. Users said they wanted the latter, as “one good book is worth ten bad ones.” Readers also felt that their recommendations served as commercial tie-ins to get them to continue reading through larger series or pick up commonly popular books that had little to do with their taste.
Additionally, users noted issues with using the search feature, stating a desire for a tag system when searching for books. They also desired a cleaner and simpler UI for both searches and recommendations.
Synthesizing user insights through affinity mapping
After gathering useful insights from Goodreads users, I synthesized this information using affinity mapping. This allowed me to group recurring themes, ideas, and pain points into categories to begin identifying patterns in user behavior. I found actionable insights in this raw information and moved forward with a clearer sense of what users are looking for from a book tracking tool.
2. Define
Framing the problem with POVs and HMWs
With a clearer understanding of the challenges Goodreads users faced, I used POVs and HMWs to frame these insights and craft a streamlined version of the users’ pain points to reference moving forward.
POV: Sandy wants a tool that gives her accurate recommendations based on her reading habits so she can find new books to read, but she doesn't trust the recommendations that Goodreads gives her.
HMW: How might we create and sustain trust with users who access the Goodreads recommendation system to help them find recommendations that are right for them?
POV: Sandy wants to view her recommended books in one simple place to save time when looking for things to read, but it's difficult to do this on Goodreads because her recommendations are scattered.
HMW: How might we centralize user recommendations on Goodreads so that we can decrease frustration and time spent in searching for new books?
Developing personas to capture the core user types
After gathering information from current Goodreads users, I synthesized their experiences into two core personas: Sandy, the Goodreads newbie who wants better recommendations and a way to connect with her local community; and Cristobal, the Goodreads veteran looking for a simpler way to search for and categorize the books he reads.
Sandy and Cristobal are the result of thoughtful research and insightful user interviews. They represent the experiences and needs of real Goodreads users who are seeking a simpler and more effective platform for their reading needs.
3. Ideate
Mapping out goals for user and business
With a strong grasp on the users needs, I decided to map out goals for both user and business, while also noting technical considerations. Through my POVs and HMWs, I determined that users required a way to receive more accurate recommendations and spend less time searching for them, and, additionally, a way to interact more with their local community. Goodreads, as a business, would seek to retain user engagement and offer a seamless experience with local library integration.
Brainstorming features and prioritization
With my personas firmly rooted in the project’s core user types, I brainstormed key features for an MVP and prioritized them according to my research. I determined that the most valuable way to synthesize user needs would be by integrating Goodreads user recommendations with local libraries. This hypothetical tool would connect Goodreads users with real librarians in their communities to get hand-picked recommendations that truly fit their reading preferences, while also connecting readers with other readers in their area.
Creating a user flow to map out the experience
With clear goals laid out, the next step was to translate them into a well-defined structure to support the user journey. Referencing my personas and feature roadmapping, I planned out a comprehensive user flow that encompassed each interaction a user might have when requesting or viewing recommendations. My goal was to present the user with a simple and intuitive system that matched the current UI and interactions of Goodreads.
4. Prototype
Sketching things out with low-fidelity wireframes
With my user flow fleshed out, I began sketching low-fidelity wireframes. Considering my persona’s needs and pain points, I started translating key moments in the user journey through link their library, requesting recommendations, and viewing completed requests. I used this time to explore different layouts for the main menu and completed recommendations pages, while also planning out the flows to add your library and request recommendations.
Desktop low-fidelity sketches
Testing low-fidelity screens for usability and design choices
The next step was to test my low-fidelity screens with Goodreads users. I conducted one-on-one user testing sessions over zoom with five participants to understand what users thought about the tasks of linking their library, requesting recommendations, and viewing completed requests. Running through these three task flows, I was received valuable information that helped me to solidify my designs.
Key findings
Users wanted an option on the request form to choose librarian they receive recommendations from.
Users suggested updates to criteria and tags on the Request Recommendations form based on star rating, editions, and other factors.
Users suggested updates to the information listed on the Recommendations page based on book statistics and other factors.
Users had valuable insights as to the UI of the Recommendations page.
Using initial sketches to create high-fidelity wireframes
After solidifying the user experience through low-fidelity wireframes, I was ready to bring the tool to life with high-fidelity wireframes. From interviews to project goals to studying Goodreads’ existing UI, this was the product of thorough research and thoughtful decision making.
Desktop high-fidelity screens
5. Test
Usability testing with high-fidelity prototypes
With the Librarian Recommendations Tool prototype ready, it was time to place the product into the hands of real users. With six Goodreads users, I conducted one-on-one usability testing sessions with my finished prototype. The goal was see how easily users could complete tasks, identify usability issues, and ensure that the tool was intuitive and easily navigable.
This final round of testing confirmed that the Librarian Recommendation Tool was on the right track. All users were able to complete each task within the given time frame with minimal misclicks. The tests also highlighted key areas for improvement, like adding a link to the tool directly on the homepage, consolidating menu options, and reconsidering the use of tags.
6. Iterate
Iterating on the prototype with high priority revisions
The usability testing phase was highly illustrative and reaffirmed many of the decisions I made throughout the design process, while also highlighting key areas to improve the user experience.
Here are a few of the iterations I made:
Added homepage navigation: before and after
Consolidated menu options to eliminate redundancies.
Reworked tag system: before and after
Added link to tool from homepage for clearer path of navigation.
Reworked Recommendations menu: before and after
Reworked tags to allow users to select their top three.
Reworked recommendation to list books in order of the number of tags included.
Learning & next steps
Designing a new tool for Goodreads wasn’t just about helping readers find better book recommendations: it was about designing the kind of tool that I’d want to use. As a voracious reader who’s always seeking the next good book, I found tremendous pleasure in talking with other readers about their experiences, and this project slowly became the culmination of our shared desire for connection through reading.
Throughout the journey, I learned to look at an real product and design for an existing audience. From studying current site metrics to adapting Goodreads’ UI and branding, I gained confidence in my ability to work within existing design systems and deliver a product that reflects core brand values while also providing a new and engaging experience.
Going forward, I’ll continue to be an active user of Goodreads, while also paying attention to any site changes that are made. Additionally, I’m moving on with a deepened interest in exploring how readers can connect with their communities and local libraries. The Librarian Recommendation Tool may have been a hypothetical, but I’m walking away with a greater passion for designing valuable experiences to aid readers everywhere.