Connecting discovery, purchase and reading around one book object
Reado is a mobile reading app where readers discover books, buy them, read them and follow other readers. The redesign reduced decision friction from discovery to purchase, made reading progress and return paths visible, and established reusable patterns across discovery, commerce, reading and community.
Summary
Objectives
- Reduce decision friction from discovery to book detail and purchase.
- Make reading progress and return paths visible across sessions.
- Establish reusable patterns for discovery, commerce, reading and community.
Role
Product Designer · Product framing, UX/UI, prototyping and system design
Scope
Discovery and book details · Reader · Library and community
Membership and checkout · Mobile design system
Key Skills
- Product Design
- Information Architecture
- Interaction Design
- Mobile Design Systems
- Conversion-focused UX
- AI Interaction Design
Impact
- Discovery, purchase and retention unified around one book object.
- Clearer entry points into reading and progress.
- A reusable mobile system for future content and premium flows.
Three product problems shaped the redesign
01 · Discovery without a clear next step. Recommendations, the daily pick and promotional modules competed for attention, while entry points into a book did not consistently expose a next action. Prioritising personalised picks and attaching a clear action to every entry point would reduce decision effort.
02 · Dense content without hierarchy. Cards, accents and sections carried similar weight, making long discovery pages hard to scan. Clear surface roles, spacing and section anatomy would separate primary content from supporting modules.
03 · Navigation state was easy to miss. The bottom navigation blended into changing page content, weakening orientation. Stronger separation and a persistent selected state would improve recognition without increasing navigation height.
Prioritise the next best reading action
I structured Discover as a decision hierarchy: personalised recommendations first, an actionable daily pick second, then trends, community signals and premium content. Consistent section anatomy and progressive disclosure keep a long feed scannable while preserving breadth.
Reduce uncertainty before the purchase decision
The first viewport brings together match rationale, format, availability, price and the two primary outcomes — buy or save. Deeper content is sequenced by decision value: description and author context first, then related titles and reviews. Readers can evaluate the book without losing the conversion path.
Protect reading focus without hiding control
I reduced the reader to persistent context — book, chapter, progress and navigation — and moved infrequent preferences into one settings layer. Highlights and notes stay attached to the text, while typography, spacing and theme controls appear only on demand.
Design return paths, not isolated features
Library and community share the same reading objects. The library prioritises continuation, goals and saved titles; community adds recommendations and discussion without creating a parallel content model. This creates multiple reasons to return while keeping reading — not social activity — the organising principle.
Make premium value and billing terms comparable
I placed benefits before plan choice, kept monthly and annual options in the same comparison, and carried the selection into a persistent order summary. Price today, trial period, renewal date and payment method remain visible before confirmation, reducing ambiguity at the highest-risk point of the flow.
Use AI to clarify intent, not replace choice
AI-assisted search extends an existing discovery task by translating a vague reading intention into visible, editable criteria. Keeping the reasoning inspectable preserves user control and makes recommendations easier to trust.
AI-assisted search is a later concept exploration created for this case.
What I learned
Design around the book, not the screen
The journey became coherent when discovery, purchase, library, reading and community were modelled around the same book object. This exposed broken hand-offs early and made the next action consistent across contexts.
Use hierarchy to manage content density
The problem was not too much content but too many elements competing at the same level. Separating primary decisions, supporting modules and passive information improved scanability without removing useful depth.
Encode decisions in reusable behaviour
The system captures interaction rules for book cards, navigation, reader controls, social content and commerce. Reuse at the behaviour level keeps new features consistent instead of only matching their visual style.