Fawesome Channels Phase 1, rebrand guide for Roku
2 Phased CTV Visual Rebrand
Fawesome’s sub-channels already performed, but on Roku’s channel store ~97% of tiles looked the same. At 10–12 ft, micro-copy is unreadable, so tiles that ask viewers to read lose to tiles that signal. We redesigned tiles to declare genre instantly using a hollow brand wordmark and genre-first visuals, shipped via a vector-first, automation-assisted pipeline so quality held across sizes. No changes to catalog, pricing or promos, this was artwork as a growth lever.
Role
Design Lead
Duration
Jan - Mar 2023
Platform
Web, Roku and Fire TV
Collaborators
July, 2024
overview.
Fawesome and its genre sub-channels already performed, but blended into a store where ~97% of tiles looked the same. Phase 1 focused on catalysing Unique Users and CTR through a visual rebrand on Roku. We prioritised the Lifestyle, Movies, and iFood channels, scaling genres in waves (P0 → P2)
Roku Channel Store screen
problem statement.
At living room distance, micro-text is unreadable and users skim in bursts of 0.5 - 1.0s. Our solid, label-heavy tiles were indistinguishable in the grid, which slowed recognition and suppressed clicks. We needed tiles that declare genre instantly and differentiate on sight, with zero changes to content or pricing.
Behavior target: faster pre-attentive recognition at 10–12 ft. Metric: CTR and Launches
pain points.
Skimming with a remote rewards contour, contrast, and category cues, it punishes dense labels and fine details. Pain points were consistent: small real estate, low contrast, noisy backgrounds, and look-alike channel posters that increased decision friction.
Viewers wanted “see → know” cognition, not “see → read”
Busy tiles + micro copy hurt legibility + inconsistent genre cues slowed scan speed.
Visual representation of pain point
Viewer's experience
user research.
We tested on real TVs (720p LCD to 4K HDR OLED) in bright and dim rooms and mapped the journey:
Store impression → Channel skim → Genre recognition → Click (CTR) → App Launch → Start Friction concentrated at “recognize genre” and “click.”
The research conclusion was to cut classification time by making the tile carry the semantic load - genre first, text later.
hypothesis.
If channel tiles signal genre visually (not textually), then CTR rises.
If we invert the store’s dominant pattern, prioritising visual and hollow typography, the tile stands out and wins attention!
If we codify spacing, stroke weights by size and contrast floors and ship via a vector-first, automated pipeline, quality will scale across sizes and channels.
Users scan in sub-second bursts. Visual+ colour is decoded faster than text at TV viewing distance.
Design for 0.3-Second Scans
design decisions.
reimagined experiences
a. working with the Art Director
Partnering with the AD, we shipped a unified hollow brand silhouette system and genre semantics that made tiles instantly scannable across dark UIs. We set safe zones by size bucket, minimum outline weights, luminance floors to prevent wordmark collapse, and banned micro-copy on tiles. I led a full visual revamp per genre, sourcing raw frames, composing focal subjects, and applying color ramps aligned to genre mood.
Visual lockup (genre identity first, copy free tiles)
Safe zones: 10% of the shortest edge per S/M/L tile bucket
Min outline weight: ≥ 2 px
Luminance floor: key visual bg ≥ 80, contrast ≥ 7:1
for wordmarks (≥4.5:1 for large icons) on dark UIs
No micro-copy on tiles, genre-specific colour ramps & mood cues
Genre art rebuild (image selection + composition)
Visual silouhette
Genre visual first, copy free - Vector specs
b. working with the Product Manager
We sequenced a P0→P2 rollout by impact × risk.
Each channel ran artwork only A/Bs for 14 days (no copy/UI changes) with a ship bar of ≥ +10-15% CTR and healthy guardrails (starts, watch time, home launch). Our PM executed same-day publishes post-approval to reduce seasonality drift and keep experiments clean.
c. working with the Content Manager
Partnering with the Content Manager, we built a genre-correct background library, migrated all descriptive copy off tiles to metadata surfaces and ASO descriptions of the channel.
The content team also analysed top-viewed and popular titles within each genre to surface audience preferences, enabling more specific channel artworks. Because each primary genre often spans multiple secondary subgenres, we prioritised creative choices around the secondary genres viewers favour.
The CM then ran seasonal refreshes through a ruleset that preserved the design system. In A/B tests, these copy-free, genre-true channel tiles contributed to the observed uplifts.
d. working with Data Analyst
I partnered with the Data Analytics team to run a channel artwork only A/B test, with matched dayparts, annotating seasonal events and title-driven spikes to de noise reads. We tracked CTR, Home-screen Launches, Search Launches, Unique Users and Watch Time as a secondary signal, with attribution checks to confirm causal lift.
The team quantified novelty decay (Weeks 1→4) and segmented outcomes by platforms Roku, Fire TV, Apple TV, Comcast and Xumo, to ensure results were durable and platform agnostic.
final verdict.
Break sameness and boost scan speed by using genre diptychs (two-panel compositions), impactful focal visuals + hollow wordmarks on Channel Store tiles for stronger peripheral vision read. Reclaim tile real estate for genre, ban micro-copy on tiles as all detail lives on metadata surfaces.
Enforce contrast and luminance floors and minimum stroke weights per export size to keep wordmarks legible on small sizes. Ship vector-first with automated exports and filename linting to prevent drift.
Operate a genre ownership model for speed and depth, governed by design tokens, TV-distance QA (~8 ft), weekly crits and a final brand gate, the same cadence that underpinned our measured uplifts.
Move the vertical slider to view comparison
implementations.
We treat implementation as a design governed experiment: vectors export through a tokenised system (stroke, luminance, contrast) with TV distance QA to ensure the tile is legible and genre-true before it ships.
For each genre, we create multiple options and route them through internal gates.
Design → Content Manager → Art Director → CEO, to shortlist the final.
Each channel then runs a Roku artwork-only A/B (14 days) with matched dayparts and annotated seasonality so artwork is the only variable, we judge outcomes on CTR as primary and home-screen launches and watch time as guardrails across device strata.
If a variant misses the ≥ +12% ship bar, we revert within and iterate on composition, crop, subject scale, and colour ramp, keeping the loop fast and the reads clean.
Multiple genre based options per channel, final selection approved by stakeholders.
Topical updates for high-priority channels drove an average +12% increase in home screen launches.
Genre-specific visual with hollow text effectively break the Roku channel store's sameness and drive clicks, bigger click lifts correlate with bigger launch lifts.
Vector-first and tokens aren’t “nice to haves”, they protect the gains at large sizes.
Top 9 Channels CTR ≥ ~78%, averaged +69.8% launch
20/37 Channels (54.1%) ≥ +50% click lift
31/37 Channels (83.8%) ≥ +30% click lift
Channel Performance Metrix on Roku
Channel* Performance Metrix on Roku
Top 9* CTR ≥ ~78%
averaged +69.8% launch
20/37* (54.1%)
≥ +50% click lift
31/37* (83.8%)
≥ +30% click lift
| Channel | % increase in Clicks | % increase in Home Screen Launch |
|---|---|---|
| Horror | 78.53% | 63.47% |
| Action | 43.30% | 37.20% |
| Sci-fi | 93.72% | 84.39% |
| Western | 32.40% | 36.87% |
| Black Cinema | 85.40% | 79.64% |
| Drama | 23.56% | 19.59% |
| Crime | 72.20% | 37.80% |
| Classic | 46.49% | 42.10% |
| Kids | 99.47% | 48.30% |
| Sports | 17.50% | 12.87% |
| Romance | 86.40% | 62.70% |
| Mystery | 43.78% | 38.50% |
| Thriller | 61.93% | 34.30% |
| Faith | 12.40% | 8.37% |
| Soccer | 19.30% | 16.90% |
| Family | 28.45% | 24.56% |
final takeaways.
The fastest way to move CTV discovery is to reduce classification time.
On Roku, Genre-First tiles + hollow wordmark did exactly that!
Users recognised faster, clicked more and launched more.
The pattern followed a familiar curve: a small early novelty spike that tapered and then stabilised at high lifts by Week 4. Hollow wordmarks broke the grid and won attention, but over time, they lost read-weight on dark UIs at living room distance, creating readability friction despite strong genre signalling.
phase 2.
To sustain recognition and stand out in noisy rows, we kept the parts that worked and changed the instrument.
By introducing a solid NEON wordmark as the focal element.
40/60 layout (≈40% type / 60% art) for balance and speed of read
High-contrast spec (safe zones, luminance floors) tuned for dark UIs at 8-10 ft
Rolled out from Movies to Lifestyle and Food, across Roku and other CTV platforms
Phase 1 won the scan. Phase 2 locks the read, preserving the recognition advantage while delivering durable legibility and sustained lifts across CTV.
neon: Why & What.
After 12 months of Hollow, we observed read-weight loss on dark UIs and quality drift at large sizes. Phase 2 retains genre-first backgrounds but transitions to solid Museo neon type to enhance living room readability and incorporates tokenised, vector-first production to maintain quality consistency as we scale.
Result: clearer read at 10–12 ft, stronger brand recall, and consistent quality across 208 channels.
Introduction of Neon color in channel names
| Area | Phase 1 (Hollow) | Phase 2 (Neon) | Why it matters |
|---|---|---|---|
| Wordmark | Outline, high contrast | Solid Museo neon | Reads faster on dark UIs, on-brand |
| Layout | Visual-dominant | 40/60 brand block : visual | Brand recall; still genre-first |
| Palette | Dark ramps behind outline | Capped-luminance neon palette | Avoids HDR bloom, keeps punch |
| Tokens | Stroke by size (S2/S3) | Stroke + luminance floors + safe zones | Consistency across sizes |
| Pipeline | Mixed raster exports | Vector-first + automated preflight | No pixelation at large sizes |
| Governance | Manual review | Linted naming, TV-distance QA, rollback in minutes | Reliable, scalable ops |
| Area | Phase 1 (Hollow) | Phase 2 (Neon) | Why it matters |
|---|---|---|---|
| Wordmark | Outline, high contrast | Solid Museo neon | Reads faster on dark UIs, on-brand |
| Layout | Visual-dominant | 40/60 brand block : visual | Brand recall; still genre-first |
| Palette | Dark ramps behind outline | Capped-luminance neon palette | Avoids HDR bloom, keeps punch |
| Tokens | Stroke by size (S2/S3) | Stroke + luminance floors + safe zones | Consistency across sizes |
| Pipeline | Mixed raster exports | Vector-first + automated preflight | No pixelation at large sizes |
| Governance | Manual review | Linted naming, TV-distance QA, rollback in minutes | Reliable, scalable ops |
palatte and tokens.
Faster recognition at 10–12 ft, Tokens: minimum x-height, 40/60 layout, luminance floors/HDR cap, automated QA → Result: clearer read and consistent launches
the upgrade.
We replaced the hollow outline with a solid Museo neon wordmark and kept the original genre backgrounds wherever they passed readability and contrast checks. When a background competed with the neon or underperformed at TV distance, we swapped in a single-subject, genre-true visual that preserved the same mood and semantics.
The story stays the same, but the signal is cleaner and faster to recognize on dark UIs, without losing identity.
Phase 1 - Hollow: Outline draws attention, loses read-weight on dark UI at 10–12 ft.
Phase 2 - Neon: Solid Museo neon in a 40/60 layout, stable read at living-room distance.
scale & operations.
Stat cards showing channel volume footprint across 3 verticals
39
Fawesome Movie Channels
14+40
Fawesome Travel + Lifestyle Channels
29
iFOOD Channels
Channel layout system by vertical - Movies, iFOOD, Travel
final verdict.
Phase 2 (matched cohort).
Shortens classification time and keep readability at TV distance.
WITH Neon solid wordmark, 40/60 layout, guardrails (safe zone 10%, contrast ≥7:1, no micro copy)
Phase 1 won attention via pattern break.
Phase 2 held performance over four weeks (novelty taper without collapse). Tiles remained legible at 8–10 ft across Roku, Fire TV, Apple TV, Xumo.
Verdict. Keep genre-first. Use neon solid to carry read-weight on dark UIs. The system scaled and cleared the ≥12% per-channel click target across the cohort.
Phase 2 converts recognition to action, intent and totals both rise
Phase 2 shows minimal week on week taper across channels, emphasizing stronger retention than Phase 1
impact.
Clicks by Channel (Before vs After). We updated all channels across three verticals, matched the cohort analysis. This chart uses the same channels shown in Phase 1.
Phase 1 captured attention.
Phase 2 converted it into action, targeting ≥12% click growth per channel.
Higher Home Screen Launches and a rightward click shift for every channel
Fawesome movie channels
This panel shows the Teen channel concept library. After testing six directions, we trialed Option 5 for a month, but it diverged from our system tokens. We standardized on Option 6 for brand consistency, neon-first system and improve recognition on dark UIs.
Food channels
topical ifood refreshes.
We built a seasonal creative system for iFood parent channel: food-first tiles tied to major holidays and produce cycles. The topical match reduced recognition time on the shelf and consistently lifted home-screen launches and watchtime at each refresh. The grid displays representative creatives, while the adjacent bars illustrate the post refresh impact.
App launch and watch time inflation with each topical change
Fawesome travel channels
Fawesome travel channels
the finish.
We locked the read and the metrics moved!
Measured impact (Phase 2, CTR lift):
Roku (medians) - Movies +30.3%, iFOOD +24.6%, Travel ≈+22.8%, Lifestyle +23.4%.
Fire TV (ratio-derived) - Movies +24.6%, iFOOD ≈+19.9%, Travel ≈+18.4%, Lifestyle ≈+19.0%.
Comcast & Apple TV - conservative, platform-scaled lifts, all positive, Movies lead, Lifestyle lowest consistent with observed behavior.
Why it worked: neon increased recognition certainty → more intentional starts (CTR) → more home-screen launches and longer sessions in high-intent cohorts. This wasn’t a paint job, it was a system.
How it scaled: tokens enforced minimum x-height, luminance floors + HDR caps, 10% safe-zones, 40/60 layout, with automated preflight and TV-distance QA, so we could ship consistently across 200+ surfaces and roll back in minutes if needed.
CTR lift (%) after the neon swap across Movies, iFOOD, Travel, Lifestyle for Roku
Fawesome Channels Phase 1, rebrand guide for Roku
2 Phased CTV Visual Rebrand
Fawesome’s sub-channels already performed, but on Roku’s channel store ~97% of tiles looked the same. At 10–12 ft, micro-copy is unreadable, so tiles that ask viewers to read lose to tiles that signal. We redesigned tiles to declare genre instantly using a hollow brand wordmark and genre-first visuals, shipped via a vector-first, automation-assisted pipeline so quality held across sizes. No changes to catalog, pricing or promos, this was artwork as a growth lever.
Role
Design Lead
Duration
Jan - Mar 2023
Platform
Web, Roku and Fire TV
Collaborators
July, 2024
overview.
Fawesome and its genre sub-channels already performed, but blended into a store where ~97% of tiles looked the same. Phase 1 focused on catalysing Unique Users and CTR through a visual rebrand on Roku. We prioritised the Lifestyle, Movies, and iFood channels, scaling genres in waves (P0 → P2)
Roku Channel Store screen
problem statement.
At living room distance, micro-text is unreadable and users skim in bursts of 0.5 - 1.0s. Our solid, label-heavy tiles were indistinguishable in the grid, which slowed recognition and suppressed clicks. We needed tiles that declare genre instantly and differentiate on sight, with zero changes to content or pricing.
Behavior target: faster pre-attentive recognition at 10–12 ft. Metric: CTR and Launches
pain points.
Skimming with a remote rewards contour, contrast, and category cues, it punishes dense labels and fine details. Pain points were consistent: small real estate, low contrast, noisy backgrounds, and look-alike channel posters that increased decision friction.
Viewers wanted “see → know” cognition, not “see → read”
Busy tiles + micro copy hurt legibility + inconsistent genre cues slowed scan speed.
Visual representation of pain point
Viewer's experience
user research.
We tested on real TVs (720p LCD to 4K HDR OLED) in bright and dim rooms and mapped the journey:
Store impression → Channel skim → Genre recognition → Click (CTR) → App Launch → Start Friction concentrated at “recognize genre” and “click.”
The research conclusion was to cut classification time by making the tile carry the semantic load - genre first, text later.
hypothesis.
If channel tiles signal genre visually (not textually), then CTR rises.
If we invert the store’s dominant pattern, prioritising visual and hollow typography, the tile stands out and wins attention!
If we codify spacing, stroke weights by size and contrast floors and ship via a vector-first, automated pipeline, quality will scale across sizes and channels.
Users scan in sub-second bursts. Visual+ colour is decoded faster than text at TV viewing distance.
Design for 0.3-Second Scans
design decisions.
reimagined experiences
a. working with the Art Director
Partnering with the AD, we shipped a unified hollow brand silhouette system and genre semantics that made tiles instantly scannable across dark UIs. We set safe zones by size bucket, minimum outline weights, luminance floors to prevent wordmark collapse, and banned micro-copy on tiles. I led a full visual revamp per genre, sourcing raw frames, composing focal subjects, and applying color ramps aligned to genre mood.
Visual lockup (genre identity first, copy free tiles)
Safe zones: 10% of the shortest edge per S/M/L tile bucket
Min outline weight: ≥ 2 px
Luminance floor: key visual bg ≥ 80, contrast ≥ 7:1
for wordmarks (≥4.5:1 for large icons) on dark UIs
No micro-copy on tiles, genre-specific colour ramps & mood cues
Genre art rebuild (image selection + composition)
Visual silouhette
Genre visual first, copy free - Vector specs
b. working with the Product Manager
We sequenced a P0→P2 rollout by impact × risk.
Each channel ran artwork only A/Bs for 14 days (no copy/UI changes) with a ship bar of ≥ +10-15% CTR and healthy guardrails (starts, watch time, home launch). Our PM executed same-day publishes post-approval to reduce seasonality drift and keep experiments clean.
c. working with the Content Manager
Partnering with the Content Manager, we built a genre-correct background library, migrated all descriptive copy off tiles to metadata surfaces and ASO descriptions of the channel.
The content team also analysed top-viewed and popular titles within each genre to surface audience preferences, enabling more specific channel artworks. Because each primary genre often spans multiple secondary subgenres, we prioritised creative choices around the secondary genres viewers favour.
The CM then ran seasonal refreshes through a ruleset that preserved the design system. In A/B tests, these copy-free, genre-true channel tiles contributed to the observed uplifts.
d. working with Data Analyst
I partnered with the Data Analytics team to run a channel artwork only A/B test, with matched dayparts, annotating seasonal events and title-driven spikes to de noise reads. We tracked CTR, Home-screen Launches, Search Launches, Unique Users and Watch Time as a secondary signal, with attribution checks to confirm causal lift.
The team quantified novelty decay (Weeks 1→4) and segmented outcomes by platforms Roku, Fire TV, Apple TV, Comcast and Xumo, to ensure results were durable and platform agnostic.
final verdict.
Break sameness and boost scan speed by using genre diptychs (two-panel compositions), impactful focal visuals + hollow wordmarks on Channel Store tiles for stronger peripheral vision read. Reclaim tile real estate for genre, ban micro-copy on tiles as all detail lives on metadata surfaces.
Enforce contrast and luminance floors and minimum stroke weights per export size to keep wordmarks legible on small sizes. Ship vector-first with automated exports and filename linting to prevent drift.
Operate a genre ownership model for speed and depth, governed by design tokens, TV-distance QA (~8 ft), weekly crits and a final brand gate, the same cadence that underpinned our measured uplifts.
Move the vertical slider to view comparison
implementations.
We treat implementation as a design governed experiment: vectors export through a tokenised system (stroke, luminance, contrast) with TV distance QA to ensure the tile is legible and genre-true before it ships.
For each genre, we create multiple options and route them through internal gates.
Design → Content Manager → Art Director → CEO, to shortlist the final.
Each channel then runs a Roku artwork-only A/B (14 days) with matched dayparts and annotated seasonality so artwork is the only variable, we judge outcomes on CTR as primary and home-screen launches and watch time as guardrails across device strata.
If a variant misses the ≥ +12% ship bar, we revert within and iterate on composition, crop, subject scale, and colour ramp, keeping the loop fast and the reads clean.
Multiple genre based options per channel, final selection approved by stakeholders.
Topical updates for high-priority channels drove an average +12% increase in home screen launches.
results.
Genre-specific visual with hollow text effectively break the Roku channel store's sameness and drive clicks, bigger click lifts correlate with bigger launch lifts.
Vector-first and tokens aren’t “nice to haves”, they protect the gains at large sizes.
Top 9 Channels CTR ≥ ~78%, averaged +69.8% launch
20/37 Channels (54.1%) ≥ +50% click lift
31/37 Channels (83.8%) ≥ +30% click lift
Channel Performance Metrix on Roku
Channel* Performance Metrix on Roku
Top 9* CTR ≥ ~78%
averaged +69.8% launch
20/37* (54.1%)
≥ +50% click lift
31/37* (83.8%)
≥ +30% click lift
| Channel | % increase in Clicks | % increase in Home Screen Launch |
|---|---|---|
| Horror | 78.53% | 63.47% |
| Action | 43.30% | 37.20% |
| Sci-fi | 93.72% | 84.39% |
| Western | 32.40% | 36.87% |
| Black Cinema | 85.40% | 79.64% |
| Drama | 23.56% | 19.59% |
| Crime | 72.20% | 37.80% |
| Classic | 46.49% | 42.10% |
| Kids | 99.47% | 48.30% |
| Sports | 17.50% | 12.87% |
| Romance | 86.40% | 62.70% |
| Mystery | 43.78% | 38.50% |
| Thriller | 61.93% | 34.30% |
| Faith | 12.40% | 8.37% |
| Soccer | 19.30% | 16.90% |
| Family | 28.45% | 24.56% |
final takeaways.
The fastest way to move CTV discovery is to reduce classification time.
On Roku, Genre-First tiles + hollow wordmark did exactly that!
Users recognised faster, clicked more and launched more.
The pattern followed a familiar curve: a small early novelty spike that tapered and then stabilised at high lifts by Week 4. Hollow wordmarks broke the grid and won attention, but over time, they lost read-weight on dark UIs at living room distance, creating readability friction despite strong genre signalling.
phase 2.
To sustain recognition and stand out in noisy rows, we kept the parts that worked and changed the instrument.
By introducing a solid NEON wordmark as the focal element.
40/60 layout (≈40% type / 60% art) for balance and speed of read
High-contrast spec (safe zones, luminance floors) tuned for dark UIs at 8-10 ft
Rolled out from Movies to Lifestyle and Food, across Roku and other CTV platforms
Phase 1 won the scan. Phase 2 locks the read, preserving the recognition advantage while delivering durable legibility and sustained lifts across CTV.
neon: Why & What.
After 12 months of Hollow, we observed read-weight loss on dark UIs and quality drift at large sizes. Phase 2 retains genre-first backgrounds but transitions to solid Museo neon type to enhance living room readability and incorporates tokenised, vector-first production to maintain quality consistency as we scale.
Result: clearer read at 10–12 ft, stronger brand recall, and consistent quality across 208 channels.
Introduction of Neon color in channel names
| Area | Phase 1 (Hollow) | Phase 2 (Neon) | Why it matters |
|---|---|---|---|
| Wordmark | Outline, high contrast | Solid Museo neon | Reads faster on dark UIs, on-brand |
| Layout | Visual-dominant | 40/60 brand block : visual | Brand recall; still genre-first |
| Palette | Dark ramps behind outline | Capped-luminance neon palette | Avoids HDR bloom, keeps punch |
| Tokens | Stroke by size (S2/S3) | Stroke + luminance floors + safe zones | Consistency across sizes |
| Pipeline | Mixed raster exports | Vector-first + automated preflight | No pixelation at large sizes |
| Governance | Manual review | Linted naming, TV-distance QA, rollback in minutes | Reliable, scalable ops |
| Area | Phase 1 (Hollow) | Phase 2 (Neon) | Why it matters |
|---|---|---|---|
| Wordmark | Outline, high contrast | Solid Museo neon | Reads faster on dark UIs, on-brand |
| Layout | Visual-dominant | 40/60 brand block : visual | Brand recall; still genre-first |
| Palette | Dark ramps behind outline | Capped-luminance neon palette | Avoids HDR bloom, keeps punch |
| Tokens | Stroke by size (S2/S3) | Stroke + luminance floors + safe zones | Consistency across sizes |
| Pipeline | Mixed raster exports | Vector-first + automated preflight | No pixelation at large sizes |
| Governance | Manual review | Linted naming, TV-distance QA, rollback in minutes | Reliable, scalable ops |
palatte and tokens.
Faster recognition at 10–12 ft, Tokens: minimum x-height, 40/60 layout, luminance floors/HDR cap, automated QA → Result: clearer read and consistent launches
the upgrade.
We replaced the hollow outline with a solid Museo neon wordmark and kept the original genre backgrounds wherever they passed readability and contrast checks. When a background competed with the neon or underperformed at TV distance, we swapped in a single-subject, genre-true visual that preserved the same mood and semantics.
The story stays the same, but the signal is cleaner and faster to recognize on dark UIs, without losing identity.
Phase 1 - Hollow: Outline draws attention, loses read-weight on dark UI at 10–12 ft.
Phase 2 - Neon: Solid Museo neon in a 40/60 layout, stable read at living-room distance.
scale & operations.
Stat cards showing channel volume footprint across 3 verticals
39
Fawesome Movie Channels
14+40
Fawesome Travel + Lifestyle Channels
29
iFOOD Channels
Channel layout system by vertical - Movies, iFOOD, Travel
final verdict.
Phase 2 (matched cohort).
Shortens classification time and keep readability at TV distance.
WITH Neon solid wordmark, 40/60 layout, guardrails (safe zone 10%, contrast ≥7:1, no micro copy)
Phase 1 won attention via pattern break.
Phase 2 held performance over four weeks (novelty taper without collapse). Tiles remained legible at 8–10 ft across Roku, Fire TV, Apple TV, Xumo.
Verdict. Keep genre-first. Use neon solid to carry read-weight on dark UIs. The system scaled and cleared the ≥12% per-channel click target across the cohort.
Phase 2 converts recognition to action, intent and totals both rise
Phase 2 shows minimal week on week taper across channels, emphasizing stronger retention than Phase 1
impact.
Clicks by Channel (Before vs After). We updated all channels across three verticals, matched the cohort analysis. This chart uses the same channels shown in Phase 1.
Phase 1 captured attention.
Phase 2 converted it into action, targeting ≥12% click growth per channel.
Higher Home Screen Launches and a rightward click shift for every channel
movies.
Fawesome movie channels
This panel shows the Teen channel concept library. After testing six directions, we trialed Option 5 for a month, but it diverged from our system tokens. We standardized on Option 6 for brand consistency, neon-first system and improve recognition on dark UIs.
ifood.tv
Food channels
topical ifood refreshes.
We built a seasonal creative system for iFood parent channel: food-first tiles tied to major holidays and produce cycles. The topical match reduced recognition time on the shelf and consistently lifted home-screen launches and watchtime at each refresh. The grid displays representative creatives, while the adjacent bars illustrate the post refresh impact.
App launch and watch time inflation with each topical change
lifestyle.
Fawesome travel channels
travel.
Fawesome travel channels
the finish.
We locked the read and the metrics moved!
Measured impact (Phase 2, CTR lift):
Roku (medians) - Movies +30.3%, iFOOD +24.6%, Travel ≈+22.8%, Lifestyle +23.4%.
Fire TV (ratio-derived) - Movies +24.6%, iFOOD ≈+19.9%, Travel ≈+18.4%, Lifestyle ≈+19.0%.
Comcast & Apple TV - conservative, platform-scaled lifts, all positive, Movies lead, Lifestyle lowest consistent with observed behavior.
Why it worked: neon increased recognition certainty → more intentional starts (CTR) → more home-screen launches and longer sessions in high-intent cohorts. This wasn’t a paint job, it was a system.
How it scaled: tokens enforced minimum x-height, luminance floors + HDR caps, 10% safe-zones, 40/60 layout, with automated preflight and TV-distance QA, so we could ship consistently across 200+ surfaces and roll back in minutes if needed.
CTR lift (%) after the neon swap across Movies, iFOOD, Travel, Lifestyle for Roku