PROJECT START

Brief

A set of images collected together, organised and activated to communicate a particular insight, message or narrative.

Duration

5 Weeks

DISCOVERY

Researching the problem space.

Overview

Exploratory research around book collections and reading to expand knowledge on the full context of the domain’s unknowns, preventing a validation exercise.

Research Goals

To avoid assumptions such as the *False-Consensus effect, I am aiming to gather data to understand...

Problem spaces which present the greatest opportunities for improvement.
The key users and other stakeholders’ use cases in context.
The current landscape of the domain, acknowledging indirect and direct competitors.

Research Methods

1. PRIMARY RESEARCH : USER INTERVIEWS

I conducted remote and in-person interviews on 5 people who were avid readers to get qualitative data. Questions were based around the research goals in a user context. Although, the most surprising discoveries were uncovered with un-rehearsed, follow-up questions.

I loosely followed the Nielson Norman Group guidelines on user interviews which also raised an awareness on the limitations of this self-reported method (Pernice, 2018).

I have yet to understand how to minimise such limitations; thus, I will also be using secondary research to improve these findings.

2. EXTERNAL SECONDARY RESEARCH

Reading academic journals around the topic to help uncover new areas based on the research goals. This method is cost- and time- effective (Azarova, 2022) and helped to refine the problem statement. However, I did not have the time nor resources to validate or critically review the limitations of each finding.

3. COMPETITOR ANALYSIS

Prior research uncovered direct (publishing and review sites etc.) and indirect (shopping and media sites etc.) applications being used currently to solve users’ needs. I conducted a design audit on the major sites to see what’s working and what’s not working.

Qualitative data was gathered, from reviews online and prior user interviews, to know the user reception. Site metrics, when available, were used to garner how the current design, specifically the features of the site, were performing.

DEFINE

Synthesising research to define the problem(s) users face.

Context

The largest and most popular platform for readers by far, Goodreads, is owned by Amazon which is leveraging data within an influential social network. Common discourse is the dissatisfaction that Goodreads has not been updated to the current needs and times for a lack of competition and thus prone to media / data gatekeeping.

Problem

Through affinity mapping, I deduced the main problems for readers:

Choosing the next book to read.
Discovering persistently high-quality recommendations.
Books are just one of many entertainment forms competing for time and attention.

These are what the users seem to value most to orient their purchase of a book.

Aims and Opportunities

There were no conscious decisions when choosing books for this collection, but the design serves to reveal the inherent biases and proclivities in the “choice.”

This project could serve as an opportunity to empower readers to curate and decolonise their experience by acknowledging more diverse reading.

Considerations

Tackling the subject of diversity is an ethical minefield – people aren’t binary with considerations to “hidden” diversity thus certain diversification efforts can be at risk of backfiring (e.g. tokenism). To be more aware of these, I looked into the industry which considers this heavily – recruitment.

MORAL DESIGN

Learning about the nature of collections (their gaps and inevitable bias), thus as a designer, wanting to show a collection which considers diversity with an element of serendipity and autonomy.

ACCESSIBLE DESIGN

Which sector understands accessibility and practices it best? Government websites. Its guidance for web regulations brings greater clarity for realising the importance of accessibility.

Improving with professor feedback from a prior project, to consider those with visual impairments, and those with visual or cognitive disabilities by using a contrast checker.

Bryony Szerkeres lecture on practical design solutions also helped me to realise the longevity of a site.

From directly working with developers in an internship, it is important to acknowledge the physical context in which a site is accessed e.g. where using adaptable grids are best practice.

Feature Priority

I used the MoSCoW Analysis as a simple and clear method to avoid *feature creep – discerning the top features to include from the “Must haves” and the “Should haves”:

Dashboard of the overall contents of the collection which updates upon filtering.
Weighted book ratings
Adding tags to search through all of the book’s contents
Only show recommended reviews from notable and relevant people.
Include contextual information.

User Personas

Having discarded prior user personas as having irrelevant information from referencing other personas on the web I re-evaluated the purpose of the method – to help empathise with users.

An updated proto-persona instead reflects on storytelling and in-depth descriptions of user context and needs to safeguard against stereotypes.

Dave is an English Literature student at University. He considers himself a hardcore reader, doing it for leisure and work.

"IT’S FRUSTRATING KNOWING I CAN ONLY READ SO MANY BOOKS IN MY LIFETIME – THAT IT’S IMPOSSIBLE TO LEARN AND GAIN KNOWLEDGE ABOUT EVERYTHING."

Environment & Context

STUDENT ACCOMMODATION

Often studies at home due to the pandemic but finds it hard to concentrate because his room is cramped and it can be quite noisy.

UNIVERsiTY

Has to do presentations on the books they are studying to a diverse group of people and does work for the newspaper club so it is important he is aware of historic and current events to engage his audience.

Goals and Needs

Relevant and highly rated books to read.
Reduce the friction to understand the context and appreciate a book without having to search further.
Ensure he is always learning through diverse content.

Hesitations and Pain Points

Inconsistent book ratings which weigh differently for each person makes choosing books difficult.
Using book covers as a process of choosing a book is inefficient.
Hard to find diverse book recommendations as he is sceptical of who polices what to spotlight.

DEVELOP

Exploring different solutions and possibilities based on the user in-sights.

User Flows

Having deduced the “what” (features), “when” is just as important. (Whitenton, 2021) This method is used to organise the systems which take place within the site so the UI would be much easier to design later on. If I wanted to fully code this site in future, this document would also be helpful.

Initial Ideas

Creating Lo-fi sketches and notes based on the scope and areas for opportunity within a set time limit. Brainstorming on paper quells perfectionism and helps to bring any idea to life – aiming for quantity rather than quality for now.

Designs and major iterations based on user feedback

Homepage

Early designs were too cramped with information which de-valued the text hierarchy and reading flow. This was solved by simply opening a new page rather than having a drop-down.
Lacking in images or visual stimulation, I improved the layout with more spacing and added colours variables to tags.

Dashboard

Considering there are more countries if one were to add more to the collection, it is not good practice to show the data within a pie chart. Thus, the data hierarchy starts from continent down to country if the user wishes to know more.
Binary tag was too simplified.
Thoughtful syntax labelled as “Bookshelf Zeitgeist”

Filters

“Diversity and Inclusion” filters seem 'judgy' and does not take into context different use cases e.g. someone who is researching a particular topic where diversity is not needed.
Rather than the filters I tried to add a spotlight section however, I would need to be the judge and curator of the definition of a minority etc. of which I am unqualified.
In the end I let the user decide the kind of representation they want in their books and authors and support any use case be it research or casual.

DEVLIVER

Prototype the most feasible and impactful solution by refining the ideas.

Homepage

1. Dashboard of the overall contents of the archive as a reference – hopefully for more intentional reading where readers can know and are aware of the power of their choice.A test would need to take place to prove/disprove this assumption.

2.Although studies found the author (their reputation and the reader’s past experience with their work) and the book cover to be major contributions to a book purchase (d’Astous et al., n/a), I aim to re-orient these behaviours The underlying value in these is that by looking at the author name, someone may discern the quality of the work (based on their past experience or what they’ve heard), this can be solved with showing the book’s overall rating. Book covers are inefficient ways to choose a book. The design takes inspiration from blind books as an unbiased way to choose a book. However, in the case that they do like the book based on that information but click further to find the context e.g. the author’s from a very typical background though they want to read someone from a different view, it is up to the user to decide for themselves, if so they can just filter for those things.

3. Quantifiable yet figuratively endless, users have the opportunity to add tags themselves as description tags are heuristic labels – variables which change depending on cultural changes. Thus, new ideas can displace old ones in the same way the reception of a book may be different 20 years on. They are also colour coded for better discernment when skim reading.

Filters

These filters allow users to classify their own requirements and the website will present books on the same topic together.

1. Slimmed down the regular filters by replacing endless book categories which require tree-like filtering used by other sites by just using the tags.

2. Option to have advanced filters via. another click of a button for less casual readers. Or, If one has a certain criteria e.g. reading from a gay author, they can filter it.

Context Page

Allow for more in depth exploration of the book beyond its face-value material.

1. A space for context can help people think more critically and make connections about what they read, appreciating the ways in which something is brought about.

2. Most review sites rate books overall but do not take into account the parts that make a book which also have different weighings depending on the person, their mood etc. (e.g. some value writing over plot). There may be a higher bar of entry, but with added reviews, the precision of ratings would be greater which seems to be a fair trade-off. The book’s total rating would take in the factors which make up the book and would be calculated the same way grades are weighted nationally.

3. Content ratings which would be averaged out per user rating. Also, being upfront about content to look out for and consider, helping people to choose critically and make their own decisions and opinions aobut what they’re getting into.

4. Recommendations from notable person(s) to search for the unsearchable.

Coding an element of the design.

With the design’s foundation on grids, I wanted to replicate flexible grids in code to better understand, as a designer, best practices so that I could communicate better with developers building the product.

PROJECT END

Reflection

There is overall positive feedback, especially from those whom are the intended users, on the design and its features.

However, a valid point of contention questions the use of the “bookshelf zeitgeist,” a part health-check and data awareness as a reference for more intentional reading – the argument advocates being more upfront at the aim to encourage readers to read more diversely rather than being on the fence.

Whose job is it to communicate data to a mass audience in an unbiased way? Journalists. I remembered a video from respected journalist Johnny Harris discussing the job. It is true, being apolitical is a political statement, he notes “objectivity is a myth” (2021) where the underlying value of that is really fairness – the aim to show all sides. I want to present readers and book lovers with a site to help them find genuinely good books in a fair way.

UPON REFLECTION, THESE THINGS MAY NEED TO HAPPEN:

The “bookshelf zeitgeist” may be more useful when filtered or when users could make their own reading list. A mentor who manages a team of data scientists (who also graduated with a degree in English Literature, making him an intended user) talked about the ways in which data could be managed through different machine learning methods (e.g. collaborative filtering, association rules etc.). Arising the question: How can users thrive off of their own data?
Rather than a book blog of sorts with either myself or a small team curating books, a volunteer-led approach to de-centralise data where users can add their own books, make edits to the context pages etc. to turn the site into a large database. Coincidentally, I was listening to a TedTalk on Wikipedia. Alot of good sites are passion and love-led by their communities. Wikipedia brings free information whilst trying to be as unbiased as possible (reasonable and fair) – a 2019 group of researchers found the more an article is edited by those of differing opinions, the more balanced and productive their contributions became. (Maher, 2022). This model of volunteers may suit this site as the *book community also seek to share and help others find good books.