Redesigning Rotten Tomatoes for the Streaming Era
A speculative redesign that transforms a legacy review aggregator into an intelligent movie decision platform. This case study documents the research, strategic thinking, and design decisions behind StarRank, a concept that moves beyond passive scores toward contextual insights, social signals, and actionable streaming data.
Click to explore full designs
Project Context
Rotten Tomatoes has been the default destination for movie reviews since 1998, serving 17+ million monthly users. Yet its core movie page has remained largely unchanged for over a decade while user expectations have been reshaped by Spotify, Netflix, and Letterboxd.
For this exercise, I chose to redesign the movie detail page using F1 (2025) as a test case. F1 is an ideal candidate: a major theatrical release with cultural relevance, strong critical reception, and a global audience with regional variation in enthusiasm. It provided enough complexity to stress-test the design system across different content scenarios.
The result is StarRank, a concept that reimagines the movie page as a decision-support tool rather than a reference document. The core question driving the redesign: how might we help users make confident viewing decisions faster, while giving enthusiasts the depth they crave?
Process
The work unfolded over three days, moving through four phases: research and analysis, strategic framing, design execution, and documentation. Each phase informed the next, though the process was iterative rather than strictly linear.
Understanding the Problem
The Gap Between Information and Insight
Rotten Tomatoes serves 17+ million monthly users, yet its core movie page has barely evolved since 2010. I conducted a heuristic analysis of the current experience, examining it through the lens of modern user expectations shaped by platforms like Spotify, Netflix, and Letterboxd.
The fundamental issue is not a lack of information. It is that the information architecture treats all data as equally important, forcing users to do the cognitive work of synthesis themselves.
Current hero: Dense header competing with promotional content
Content section: Information overload without clear hierarchy
Heuristic Analysis
The Score Paradox
The Tomatometer and Audience Score are prominently displayed, yet they answer the wrong question. Users do not just want to know if a movie is good. They want to know if it is good for them.
An 82% Tomatometer tells you critics approved, but it does not distinguish between universal acclaim and polarizing opinions. A blockbuster action film and a slow-burn indie drama can have identical scores while serving completely different audiences.
Layer context onto raw scores: demographic breakdowns, sentiment analysis, and comparative positioning against similar films.
The Discovery Dead End
After viewing a movie page, users have limited pathways forward. The 'More Like This' section uses basic genre matching, and there is no connection to the user's personal taste graph.
Platforms like Spotify have trained users to expect personalization. Netflix's 98% Match indicators set expectations for relevance-aware recommendations. RT's static suggestions feel outdated by comparison.
Introduce social proof through friends' activity, surface community discussions, and position films within the context of a user's viewing history.
The Streaming Scavenger Hunt
'Where to Watch' information exists but requires multiple clicks to access. Users must leave the page to compare prices across platforms.
In 2025, the streaming landscape is fragmented across 200+ services. The friction of finding where to watch directly impacts conversion. Users who cannot immediately act on intent often abandon the decision entirely.
Elevate streaming availability to a first-class citizen. Show subscription services, rental prices, and theater showtimes in a single, scannable module.
The Passive Consumer Model
RT treats users as passive recipients of critic and audience opinions. There is no sense of community, no live discourse, and limited ways to engage beyond submitting a star rating.
Letterboxd's growth (100M+ logged films in 2024) proves that movie enthusiasts want social features. They want to share opinions, discover through friends, and participate in cultural conversations.
Transform the page from a reference document into a living community hub with discussions, prediction markets, and real-time social signals.
User Journey Analysis
Mapping the decision-making flow reveals where users exit the RT ecosystem. Each exit point represents lost engagement and eroded platform authority.
Current State: Fragmented Journey
User Intent
"Should I watch F1?"
Rotten Tomatoes
Check score: 82%
IMDb
Cross-reference: 7.6
Letterboxd
Friend reviews
JustWatch
Find streaming
Decision
15+ min elapsed
User leaves RT ecosystem 3 times before decision
Each context switch increases cognitive load and friction
Proposed State: Unified Journey
User Intent
"Should I watch F1?"
Aggregated Scores
Demographics
Friends Avg
Streaming
Global Data
StarRank
All signals consolidated in one view
Decision
2-3 min
Complete decision journey within StarRank
Reduced friction through data synthesis
The redesign collapses a fragmented, multi-platform journey into a single coherent experience. By synthesizing data from multiple sources and presenting it through the lens of user intent, StarRank transforms the movie page from a reference document into a decision engine.
The Solution
StarRank: From Aggregator to Intelligence Platform
The core insight driving this redesign is that users do not visit movie pages to consume data. They visit to make a decision. Every element should either help them decide whether to watch, how to watch, or deepen their understanding of something they have already seen.
StarRank repositions the movie page as a decision-support tool rather than a reference document. It synthesizes data from multiple sources, surfaces social signals, and provides the contextual layers that raw scores cannot. The name itself signals the shift: where Rotten Tomatoes focuses on critical consensus (fresh vs. rotten), StarRank emphasizes ranking, comparison, and positioning within a broader landscape of content.
Strategic Framework
The redesign is built on three interconnected principles that address the identified pain points while creating opportunities for differentiation:
1. Contextual Intelligence
Raw scores become meaningful when placed in context. Rather than showing 82%, show that this film ranks in the top 15% of 2025 releases, or that it outperforms 92% of racing films. Comparative framing helps users calibrate expectations.
Impact: Reduces score anxiety and builds confidence in decision-making.
2. Social Proof Integration
Integrate signals from the user's social graph: friends who watched it, their ratings, and whether they are currently watching. This transforms anonymous crowd scores into trusted recommendations from known sources.
Impact: Increases trust and creates emotional connection to content.
3. Actionable Proximity
Every piece of information should be within one click of action. Streaming availability, showtimes, and watchlist functions are elevated from buried features to persistent, accessible modules.
Impact: Reduces friction between intent and conversion.
Design Pillars
Six principles guided the interface design, each addressing specific user needs identified in the problem analysis:
- Data with Depth. Scores are entry points, not endpoints. Layer demographic breakdowns, sentiment analysis, and vibe metrics to help users understand not just what people think, but who thinks it and why.
- Global Perspective. Films resonate differently across cultures. F1 racing has deeper roots in Europe than North America. Visualizing regional reception helps users contextualize scores within their own cultural frame.
- Social Layer. Move beyond anonymous aggregation. Show friends' activity, enable discussions, and surface community insights. Transform passive consumption into participatory engagement.
- Live Elements. Static pages feel stale. Introduce prediction markets for awards, trending discussions, and real-time viewing activity to make the page feel alive and connected to the cultural moment.
- Decision Architecture. Structure information flow around the decision journey: Should I watch this? Where can I watch it? What should I know before watching? Each section answers a specific question.
- Progressive Disclosure. Surface key insights immediately, but allow users to drill deeper. The casual browser and the film enthusiast should both find the page useful without overwhelming either.
Design Walkthrough
Anatomy of the Redesign
Each section of the interface was designed to answer a specific user question. The information architecture flows from broad context (should I watch?) to specific details (production credits), following the natural decision-making progression.
Cinematic Hero with Multi-Platform Scores
The hero section establishes visual identity through a full-bleed cinematic image, immediately communicating the film's tone and production value. Below the fold, I aggregated scores from five platforms: Tomatometer, Audience Score, IMDb, Metacritic, and Letterboxd.
Design Rationale
Users currently tab between 3-4 sites to triangulate quality signals. By consolidating these upfront, we eliminate context-switching friction. The side-by-side display also surfaces interesting patterns: a film with high critic scores but low audience scores tells a different story than universal acclaim.
Key Decisions
- •Full-bleed image with gradient overlay creates premium theatrical feel
- •Movie poster anchored left provides recognizable thumbnail for sharing
- •IN THEATERS badge with director credit surfaces key decision factors immediately
- •Action buttons (Play Trailer, Watchlist) prioritized over passive reading
Consolidates the multi-tab comparison workflow into a single scannable view
Hero section with multi-platform score aggregation
Storyline, Cast & Streaming Sidebar
This section pairs narrative context (synopsis, credits) with actionable information (streaming availability). The Where to Watch sidebar persists on scroll, ensuring users can act on intent at any moment. Cast is displayed in a 2x3 grid with circular avatars, prioritizing recognition over comprehensiveness.
Design Rationale
The streaming landscape has become the primary friction point in movie discovery. By elevating availability to a persistent sidebar, we acknowledge that 'where can I watch this' often supersedes 'is this good'. The cast grid humanizes the page while avoiding the horizontal scroll pattern that hides content on desktop.
Key Decisions
- •Sticky sidebar ensures streaming CTAs remain visible during scroll
- •Subscription services separated from rent/buy options reduces cognitive load
- •Green 'Included with subscription' label signals value immediately
- •Find Showtimes CTA in red maintains hierarchy with theatrical releases
Streaming availability elevated to persistent viewport position, always accessible
Storyline and streaming sidebar
Audience Analytics & Social Signals
This section represents the most significant departure from traditional review aggregators. Age Distribution shows how different demographics respond (93% positive among 18-24). Vibe Check quantifies subjective qualities: Adrenaline Rush 97%, Visually Stunning 96%, Emotional Depth 74%. The Friends Activity sidebar surfaces your social graph's relationship with the film.
Design Rationale
Aggregate scores flatten nuance. A film that excites young viewers but leaves older audiences cold is fundamentally different from one with universal appeal. The Vibe Check addresses the 'what kind of good' question that percentages cannot answer. Social signals add trusted voices to anonymous crowds.
Key Decisions
- •Age distribution uses horizontal bars for easy comparison across groups
- •Vibe metrics use red progress bars to convey intensity and energy
- •Friends Activity shows avatars for recognition, star ratings for quick scanning
- •Award Markets sidebar adds speculative engagement for enthusiasts
Surfaces personal relevance signals that raw aggregate scores cannot convey
Audience analytics and social features
Global Reception & Sentiment Analysis
A dot-matrix world map visualizes reception heat by region, with detailed breakdowns for major markets. Europe leads at 94% (F1's spiritual home), while Asia Pacific trails at 88%. The Review Sentiment bar shows 72% positive, 20% mixed, 8% negative. Most Mentioned tags surface recurring themes: cinematography, racing sequences, sound design, IMAX.
Design Rationale
Cultural context matters enormously for films tied to specific traditions. F1 racing is religion in the UK and Europe but niche in North America. Making this visible helps users understand why scores might differ from their personal experience. The sentiment breakdown and keyword extraction save users from reading dozens of reviews.
Key Decisions
- •Dot-matrix map aesthetic matches the data-forward brand identity
- •Country flags provide instant recognition for top markets
- •Delta indicators (+5 vs avg) contextualize regional scores against baseline
- •Tag chips for keywords enable potential filtering in future iterations
Transforms geographic and sentiment data into visual patterns for rapid comprehension
Global reception map and sentiment analysis
Comparative Context & Community Discussions
How It Compares cards position the film against meaningful cohorts: +18% vs racing films, #2 vs Kosinski films, Top 15% vs 2025 releases. Hot Discussions surfaces community conversation with engagement metrics. More Like This recommends films with dual critic/audience scores for calibration.
Design Rationale
Humans are comparative thinkers. Knowing F1 outperforms 92% of racing films tells you more than an 82% score in isolation. The discussions feature transforms the page from monologue to dialogue, creating stickiness through community. Similar films with visible scores help users calibrate expectations.
Key Decisions
- •Four comparison cards answer different framing questions
- •Discussion threads show upvotes and replies to surface quality content
- •Movie posters in More Like This enable recognition-based browsing
- •Dual scores on recommendations reinforce the multi-signal philosophy
Comparative framing provides calibration that isolated percentages cannot offer
Comparison cards and community discussions
Production Credits & Platform Footer
Production details are organized in a four-column grid: Director, Producer, Screenwriter, Distributor in row one; Production Co, Rating, Genre, Language in row two; Release dates and box office in row three. The footer establishes StarRank's brand identity with navigation grouped by user intent.
Design Rationale
Reference information should be scannable but not prominent. The grid layout treats metadata with appropriate visual weight: important enough to include, not important enough to dominate. The footer copy 'The most comprehensive movie intelligence platform' reinforces the strategic positioning.
Key Decisions
- •Four-column grid maximizes information density without clutter
- •Label/value pairs maintain consistent vertical rhythm
- •Footer navigation grouped by intent: Explore, Markets, Company
- •Minimal footer respects that users rarely engage with this section
Production metadata consolidated into scannable grid, respecting its secondary importance
Production credits grid and footer
Iconography: The StarRank Badge System
One of the most recognizable elements of Rotten Tomatoes is its iconography: the red tomato, the green splat, the popcorn bucket. These symbols have become cultural shorthand for movie quality. For StarRank, I wanted to create a badge system that could achieve similar recognition while offering more nuance than a binary fresh/rotten distinction.
I designed a complete 10-badge system: five tiers for Critics scores and five for Audience scores. Each badge needed to communicate its rating at a glance while maintaining a cohesive visual language across the spectrum. The star character serves as the central mascot, with its expression, pose, color temperature, and surrounding elements all encoding quality signals.
Critics Score Badges
The critics badges use a formal, award-show aesthetic with laurel wreaths and metallic finishes. Color progresses from deep crimson (negative) through grayscale (neutral) to warm gold (positive), mirroring traditional film criticism iconography.
1/5
Critically Panned
2/5
Mixed Reviews
3/5
Worth Watching
4/5
Highly Recommended
5/5
Critics Choice
Audience Score Badges
The audience badges adopt a more playful, theatrical tone with popcorn, 3D glasses, and cinema motifs. The expressions are more animated to reflect the emotional, gut-reaction nature of audience response versus critical analysis.
1/5
Total Miss
2/5
Not Great
3/5
Decent Watch
4/5
Big Crowd Pleaser
5/5
Fan Favorite
Design System Analysis
Color as Information
The badge system uses color temperature as a primary encoding mechanism. Negative ratings employ cool, desaturated tones (grays, muted browns) or aggressive warm tones (angry reds), while positive ratings use warm metallics (gold, bronze) that evoke achievement and value. This allows users to assess quality before reading any text.
Character Expression as Shorthand
The star mascot's face changes dramatically across the scale: tears and distress at the low end, neutral contemplation in the middle, and joy at the top. These expressions leverage our innate ability to read faces, making the rating immediately intuitive even for first-time users unfamiliar with the system.
Contextual Symbolism
Each badge incorporates contextually appropriate symbols. The critics badges use laurel wreaths (awards), trophies, and crowns to reference the institutional nature of professional criticism. The audience badges use popcorn, drinks, 3D glasses, and movie tickets to ground the rating in the theatrical experience. Negative badges show spilled popcorn and splattered tomatoes, while positive ones show overflowing abundance.
Differentiation from Rotten Tomatoes
While RT uses a binary system (fresh/rotten) that loses nuance, StarRank's five-tier approach captures the full spectrum of reception. A 60% "fresh" film and a 95% "fresh" film look identical on RT. Here, they would display meaningfully different badges, giving users more accurate expectations.
Critical Analysis
Design Decisions & Trade-offs
Every design decision involves trade-offs. This section documents the key choices I made during this redesign, the alternatives I considered, and the reasoning that led to the final direction. My goal is to show not just what I designed, but how I think through complex product problems.
Decision Framework Applied
Theme
Dark
Light / Hybrid
Scores
Aggregated
RT-only
Density
Data-rich
Minimal
Social
Integrated
Utility-only
Layout
Sidebar
Linear
User Need
What problem does this solve?
Business Value
How does this differentiate?
Feasibility
What are the constraints?
Trade-off
What do we sacrifice?
1. Dark Theme vs. Light Theme
The decision: I chose a dark, cinema-inspired interface over the white backgrounds typical of review sites.
Alternatives considered: A light theme would have been safer and more accessible by default. It would also maintain closer parity with RT's existing brand. A hybrid approach with user-selectable themes was another option.
Why I chose this: The dark theme serves a strategic purpose beyond aesthetics. Movie discovery is often an evening activity, and dark interfaces reduce eye strain in low-light environments. More importantly, the dark canvas makes the cinematic imagery pop, creating an emotional resonance that a white background would flatten. The goal was to make browsing feel like the beginning of a movie experience, not like reading a Wikipedia article.
Trade-off acknowledged: Dark themes can present accessibility challenges for users with certain visual impairments. A production implementation would include a high-contrast mode and respect system preferences.
2. Aggregated Scores vs. Single Source
The decision: I display scores from multiple platforms (Tomatometer, Popcornmeter, IMDb, Metacritic, Letterboxd) simultaneously rather than privileging RT's own metrics.
Alternatives considered: Keeping focus solely on RT's proprietary scores would reinforce brand value and simplify the interface. Another option was a tabbed approach showing one score at a time.
Why I chose this: User research consistently shows that people already cross-reference multiple sites before deciding what to watch. By acknowledging this behavior and serving it directly, StarRank becomes the single destination rather than one of many tabs. This is a counterintuitive business decision: by featuring competitor scores, we actually increase user trust and session time. The implicit message is "we're confident enough to show you everything."
Trade-off acknowledged: This approach requires partnerships or API access with other platforms. It also creates dependency on third-party data availability and accuracy.
3. Data Density vs. Simplicity
The decision: I added substantial new data layers (age demographics, vibe metrics, global reception maps, prediction markets) that don't exist on the current RT site.
Alternatives considered: A minimal approach would strip the page down to essentials: poster, score, synopsis, watch options. This is the direction Netflix and Apple TV+ have taken with their in-app movie pages.
Why I chose this: The minimal approach works when you've already decided to watch something. But RT's core use case is the decision phase, where users are actively evaluating options. In this context, more information (presented well) reduces decision anxiety. The key insight is that data density and usability aren't opposed if you structure information around user questions: "Is this for me?" "What will I feel?" "Do people like me enjoy this?"
Trade-off acknowledged: More features means more to maintain, more edge cases, and more potential for cognitive overload. Progressive disclosure and careful visual hierarchy are essential to prevent the page from feeling overwhelming.
4. Social Features vs. Pure Utility
The decision: I introduced social elements (friends' ratings, community discussions, shared watchlists) that transform a reference tool into a social platform.
Alternatives considered: Staying purely utilitarian would avoid the complexity and moderation challenges of user-generated content. It would also sidestep the cold-start problem of social features that only work with critical mass.
Why I chose this: Letterboxd's success demonstrates that movie enthusiasts want community. More practically, social features transform usage patterns: instead of visiting only when you need to check a score, users return to see what friends are watching, engage in discussions, and share opinions. This increases engagement frequency and duration dramatically. The "Friends Activity" sidebar specifically addresses the trust gap: a score from a stranger means less than knowing three friends all rated it 4+ stars.
Trade-off acknowledged: Social features require significant investment in trust and safety, content moderation, and onboarding. They also create chicken-and-egg growth challenges. A phased rollout would be essential.
5. Sidebar Architecture vs. Linear Scroll
The decision: I used a persistent sidebar for "Where to Watch" and secondary content rather than placing everything in a single-column layout.
Alternatives considered: A fully linear layout would be simpler to implement and translate more cleanly to mobile. It would also avoid the potential for sidebar content to be missed.
Why I chose this: The sidebar solves a specific problem: streaming availability is the number one piece of information users need after deciding they're interested in a film. By keeping it persistently visible, I eliminate the need to scroll or hunt for this information. The sidebar also creates clear content hierarchy: primary content (the film) on the left, secondary actions (how to watch, social proof) on the right. This maps to the natural left-to-right reading flow of decision-making: evaluate, then act.
Trade-off acknowledged: Sidebars require careful responsive handling. On mobile, this content would need to be repositioned, potentially as a sticky bottom bar or integrated into the main flow with smart prioritization.
Principles That Emerged
Through this process, I arrived at a set of principles that could guide future development of the platform. These aren't rules I started with, but patterns that revealed themselves as I worked through specific design challenges:
- Serve the question, not the data.Every interface element should answer a specific user question. If you can't articulate what question a feature answers, it probably shouldn't exist.
- Confidence over completeness.Users don't need every piece of information. They need enough information to feel confident in their decision. Sometimes that's more, sometimes less.
- Context is content. A 97% audience score means nothing in isolation. It means everything when you can see that your demographic loves it, your friends rated it highly, and it's outperforming similar films.
- Make the action obvious. Every screen should have a clear next step. On a movie page, that's either "watch it" or "explore more." The design should make both paths effortless.
- Emotion is data. How a movie makes you feel is as important as whether critics think it's technically good. Design systems should capture and communicate emotional qualities, not just quantitative scores.
Reflections
Process Insights & Learnings
The tension between innovation and familiarity
Early explorations pushed further from RT's mental model: tabbed interfaces, card-based browsing, timeline views. User testing instincts suggested these would increase learning curve without proportional benefit. The final design innovates within recognizable patterns: the page still scrolls vertically, scores still appear near the top, and streaming info still lives in a sidebar. Innovation is in the data layers, not the navigation paradigm.
Data visualization requires editorial judgment
With access to demographic data, global reception, sentiment analysis, and social signals, the temptation was to show everything. Each iteration forced prioritization: What question does this answer? Who is asking? Early versions included viewing time distributions, rewatch rates, and second-weekend drops. These were interesting but did not serve the core decision. Cutting features was harder than adding them.
Dark themes carry emotional weight
The shift from RT's light theme to a dark interface was deliberate. Dark themes create a theatrical atmosphere appropriate for cinema. But they also risk feeling heavy or inaccessible. The solution was careful use of the red accent for energy, generous whitespace to prevent claustrophobia, and consistent contrast ratios to maintain readability. The result feels premium rather than oppressive.
Social features require trust architecture
Features like Friends Activity and Hot Discussions assume a logged-in state and populated social graph. In practice, these features would need graceful degradation for anonymous users and cold-start strategies for new accounts. The design assumes an ideal state; production would require careful consideration of the empty state journey.
If I Had More Time
- Mobile-first responsive design: The current design prioritizes desktop. A production implementation would need to address touch targets, collapsed sidebars, and vertical stacking for the analytics sections.
- Edge cases and error states: What happens when a film has no streaming availability? When Friends Activity is empty? When regional data is incomplete? These states would require careful design.
- Animation and micro-interactions: The static mockups do not convey how charts would animate in, how hover states would reveal additional information, or how the page would feel in motion.
- User research validation: The information hierarchy reflects assumptions about user priorities. Usability testing would validate or challenge these assumptions.
Why This Work is Relevant to Prelim
On the surface, a movie review site and banking onboarding software seem unrelated. But the core design challenge is identical: helping users navigate complex decisions with confidence.
Domain Translation Schematic
Progressive Disclosure
Show essentials first, details on demand
Contextual Framing
Numbers mean more with comparison
Action Proximity
CTAs visible when intent forms
Trust Architecture
Layer signals to build confidence