From Concept to #1: Text 02 Video Effect
CASE STUDY
VOCHI's effects worked for VOCHI's audience – bold, high-contrast, visually assertive. After Pinterest acquired VOCHI, the brief changed. Pinterest's key content categories are fashion, beauty, food, and home and in those verticals the content is the point: a styled dish, a room renovation or an outfit. An effect that fights the subject for attention doesn't support that content – it works against it. And dialing existing effects back doesn't solve it: a subtle version of an attention-grabbing effect still competes with the subject, just more quietly. Content-first required a different spatial logic entirely.
Text 02 was my concept from a two-week sprint: user-entered text placed behind the masked subject. The subject stays in front. The text fills the negative space. Expressive, but the content always wins. It shipped in roughly four weeks and became the #1 iOS effect by saves – surpassing Butterflies, which had held that spot for 1.5 years.
ROLE
Product Designer
TEAM
VOCHI
TIMELINE
2022
TOOLS
Product context
The tempting fix was to keep the same effect logic and dial things back. That doesn't solve the problem. A subtle version of an attention-grabbing effect still competes with the subject -- the competition is just quieter. Content-first effects needed a different spatial relationship between effect and subject, not a lower-intensity version of the same one.
The two-week sprint I participated in was one of the first structured attempts to figure out what that looked like in practice. I explored multiple directions in parallel and proposed several concepts; leadership selected Text 02 as one of the first to implement.
Problem
Most “bold” effects steal attention from the subject – which clashes with Pinterest content, where the object is the point. We needed something expressive but restrained: an effect that adds meaning and style without overpowering the content.
Constraints
- Must work across multiple content types and verticals
- Must be implementable fast (simple, reliable logic)
- Must keep the subject readable / central
- Must tolerate imperfect segmentation masks
- Needs adjustable settings without confusing users
some of my other early concepts
Process
After leadership selected the concept, I partnered with an engineer to define the behavior end-to-end and ship it. A lot of the real work happened in that phase -- the things that aren't visible in the final effect: what the text layer does by default, what the right range for size and opacity is, how the animation behaves, and how the effect handles edge cases like very long user input, extreme aspect ratios, or low background contrast.
The hardest part wasn't the effect itself -- it was the defaults. Even though Text 02's spatial logic is forgiving of rough masks by construction, how forgiving depends entirely on the numbers. Text that was too large or too opaque made minor mask imperfections more visible, not less; too subtle and the effect disappeared. Finding the range that held up across a wide variance of content -- different lighting, subjects, segmentation quality -- without requiring per-video tuning took real iteration.
The settings scope required the same kind of judgment. Too many adjustable parameters and users either get overwhelmed or make choices that produce results that look off. I went through several versions before landing on the right balance: enough control to feel personal, not so much that a good default result requires taste to achieve. Some parameters I initially exposed got pulled because they produced too much variance in output quality across content types.
QA was hands-on: I tested across a wide range of subjects, lighting conditions, mask qualities, and aspect ratios, and verified that settings labels, ranges, and defaults were legible to someone encountering the effect for the first time. An effect that holds up in one content type but breaks in another isn't shipped; it's a demo.
in-app experience
Impact
Text 02 reached #1 by iOS effect saves (effect_saved, ~5.68% share), surpassing Butterflies –the effect that had led for roughly 1.5 years. Early repeat usage was strong: the weekly rate rose from 23.8% at launch to a 36.1% peak in one of the first tracked weeks.
Repeat usage is the signal that matters most in a creative tool. It means the effect has become part of how people actually make content – not something they try once, find interesting, and move on from. That rate, early in the rollout, suggested Text 02 had found a real place in how users were making videos.
The concept-to-ship timeline was around four weeks total: two weeks of exploration where the concept was selected, then roughly two more weeks of implementation and QA with the engineer.
Text 02 is one of those rare ideas that stay simple from pitch to production – not because it required no judgment, but because the core concept was structurally sound enough that every downstream decision followed from it.
exported from VOCHI (no post-edit)
organic social video (TikTok / Instagram) featuring the Text 02 effect
other projects
Celebration System
→
Motion Foundations
→
connect
From Concept to #1: Text 02 Video Effect
CASE STUDY
VOCHI's effects worked for VOCHI's audience – bold, high-contrast, visually assertive. After Pinterest acquired VOCHI, the brief changed. Pinterest's key content categories are fashion, beauty, food, and home and in those verticals the content is the point: a styled dish, a room renovation or an outfit. An effect that fights the subject for attention doesn't support that content – it works against it. And dialing existing effects back doesn't solve it: a subtle version of an attention-grabbing effect still competes with the subject, just more quietly. Content-first required a different spatial logic entirely.
Text 02 was my concept from a two-week sprint: user-entered text placed behind the masked subject. The subject stays in front. The text fills the negative space. Expressive, but the content always wins. It shipped in roughly four weeks and became the #1 iOS effect by saves – surpassing Butterflies, which had held that spot for 1.5 years.
ROLE
Product Designer
TEAM
VOCHI
TIMELINE
2022
TOOLS
Product context
The tempting fix was to keep the same effect logic and dial things back. That doesn't solve the problem. A subtle version of an attention-grabbing effect still competes with the subject -- the competition is just quieter. Content-first effects needed a different spatial relationship between effect and subject, not a lower-intensity version of the same one.
The two-week sprint I participated in was one of the first structured attempts to figure out what that looked like in practice. I explored multiple directions in parallel and proposed several concepts; leadership selected Text 02 as one of the first to implement.
Problem
Most “bold” effects steal attention from the subject – which clashes with Pinterest content, where the object is the point. We needed something expressive but restrained: an effect that adds meaning and style without overpowering the content.
Constraints
- Must work across multiple content types and verticals
- Must be implementable fast (simple, reliable logic)
- Must keep the subject readable / central
- Must tolerate imperfect segmentation masks
- Needs adjustable settings without confusing users
some of my other early concepts
Process
After leadership selected the concept, I partnered with an engineer to define the behavior end-to-end and ship it. A lot of the real work happened in that phase -- the things that aren't visible in the final effect: what the text layer does by default, what the right range for size and opacity is, how the animation behaves, and how the effect handles edge cases like very long user input, extreme aspect ratios, or low background contrast.
The hardest part wasn't the effect itself -- it was the defaults. Even though Text 02's spatial logic is forgiving of rough masks by construction, how forgiving depends entirely on the numbers. Text that was too large or too opaque made minor mask imperfections more visible, not less; too subtle and the effect disappeared. Finding the range that held up across a wide variance of content -- different lighting, subjects, segmentation quality -- without requiring per-video tuning took real iteration.
The settings scope required the same kind of judgment. Too many adjustable parameters and users either get overwhelmed or make choices that produce results that look off. I went through several versions before landing on the right balance: enough control to feel personal, not so much that a good default result requires taste to achieve. Some parameters I initially exposed got pulled because they produced too much variance in output quality across content types.
QA was hands-on: I tested across a wide range of subjects, lighting conditions, mask qualities, and aspect ratios, and verified that settings labels, ranges, and defaults were legible to someone encountering the effect for the first time. An effect that holds up in one content type but breaks in another isn't shipped; it's a demo.
in-app experience
Impact
Text 02 reached #1 by iOS effect saves (effect_saved, ~5.68% share), surpassing Butterflies –the effect that had led for roughly 1.5 years. Early repeat usage was strong: the weekly rate rose from 23.8% at launch to a 36.1% peak in one of the first tracked weeks.
Repeat usage is the signal that matters most in a creative tool. It means the effect has become part of how people actually make content – not something they try once, find interesting, and move on from. That rate, early in the rollout, suggested Text 02 had found a real place in how users were making videos.
The concept-to-ship timeline was around four weeks total: two weeks of exploration where the concept was selected, then roughly two more weeks of implementation and QA with the engineer.
Text 02 is one of those rare ideas that stay simple from pitch to production – not because it required no judgment, but because the core concept was structurally sound enough that every downstream decision followed from it.
exported from VOCHI (no post-edit)
organic social video (TikTok / Instagram) featuring the Text 02 effect
other projects
Celebration System
→
Motion Foundations
→
connect
CASE STUDY
From Concept to #1: Text 02 Video Effect
VOCHI's effects worked for VOCHI's audience – bold, high-contrast, visually assertive. After Pinterest acquired VOCHI, the brief changed. Pinterest's key content categories are fashion, beauty, food, and home and in those verticals the content is the point: a styled dish, a room renovation or an outfit. An effect that fights the subject for attention doesn't support that content – it works against it. And dialing existing effects back doesn't solve it: a subtle version of an attention-grabbing effect still competes with the subject, just more quietly. Content-first required a different spatial logic entirely.
Text 02 was my concept from a two-week sprint: user-entered text placed behind the masked subject. The subject stays in front. The text fills the negative space. Expressive, but the content always wins. It shipped in roughly four weeks and became the #1 iOS effect by saves – surpassing Butterflies, which had held that spot for 1.5 years.
ROLE
Product Designer
TEAM
VOCHI
TIMELINE
2022
TOOLS
Product context
The tempting fix was to keep the same effect logic and dial things back. That doesn't solve the problem. A subtle version of an attention-grabbing effect still competes with the subject -- the competition is just quieter. Content-first effects needed a different spatial relationship between effect and subject, not a lower-intensity version of the same one.
The two-week sprint I participated in was one of the first structured attempts to figure out what that looked like in practice. I explored multiple directions in parallel and proposed several concepts; leadership selected Text 02 as one of the first to implement.
Problem
Most “bold” effects steal attention from the subject – which clashes with Pinterest content, where the object is the point. We needed something expressive but restrained: an effect that adds meaning and style without overpowering the content.
Constraints
- Must work across multiple content types and verticals
- Must be implementable fast (simple, reliable logic)
- Must keep the subject readable / central
- Must tolerate imperfect segmentation masks
- Needs adjustable settings without confusing users
some of my other early concepts
Process
After leadership selected the concept, I partnered with an engineer to define the behavior end-to-end and ship it. A lot of the real work happened in that phase -- the things that aren't visible in the final effect: what the text layer does by default, what the right range for size and opacity is, how the animation behaves, and how the effect handles edge cases like very long user input, extreme aspect ratios, or low background contrast.
The hardest part wasn't the effect itself -- it was the defaults. Even though Text 02's spatial logic is forgiving of rough masks by construction, how forgiving depends entirely on the numbers. Text that was too large or too opaque made minor mask imperfections more visible, not less; too subtle and the effect disappeared. Finding the range that held up across a wide variance of content -- different lighting, subjects, segmentation quality -- without requiring per-video tuning took real iteration.
The settings scope required the same kind of judgment. Too many adjustable parameters and users either get overwhelmed or make choices that produce results that look off. I went through several versions before landing on the right balance: enough control to feel personal, not so much that a good default result requires taste to achieve. Some parameters I initially exposed got pulled because they produced too much variance in output quality across content types.
QA was hands-on: I tested across a wide range of subjects, lighting conditions, mask qualities, and aspect ratios, and verified that settings labels, ranges, and defaults were legible to someone encountering the effect for the first time. An effect that holds up in one content type but breaks in another isn't shipped; it's a demo.
in-app experience
Impact
Text 02 reached #1 by iOS effect saves (effect_saved, ~5.68% share), surpassing Butterflies –the effect that had led for roughly 1.5 years. Early repeat usage was strong: the weekly rate rose from 23.8% at launch to a 36.1% peak in one of the first tracked weeks.
Repeat usage is the signal that matters most in a creative tool. It means the effect has become part of how people actually make content – not something they try once, find interesting, and move on from. That rate, early in the rollout, suggested Text 02 had found a real place in how users were making videos.
The concept-to-ship timeline was around four weeks total: two weeks of exploration where the concept was selected, then roughly two more weeks of implementation and QA with the engineer.
Text 02 is one of those rare ideas that stay simple from pitch to production – not because it required no judgment, but because the core concept was structurally sound enough that every downstream decision followed from it.
exported from VOCHI (no post-edit)
organic social video (TikTok / Instagram) featuring the Text 02 effect
other projects
Celebration System
→
Motion Foundations
→
connect