Canvas Health - Lucille Trepanier

Overview

Canvas Health - reducing digital waste through system-level design

Canvas Health rethinks how design tools manage invisible accumulation. By making duplication visible at the moment it happens, introducing lightweight interventions, and automating cleanup over time, the system helps designers maintain performant files while reducing unnecessary storage and energy use.

Rather than relying on manual cleanup, the product shifts behavior through feedback, nudges, and system defaults - aligning user workflows with both performance and sustainability goals.

100M+ users in Figma's global collaborative environment

$100M+ annual cloud costs from storage and sync

The Problem

Digital accumulation is invisible, so behavior never changes

Designers are not careless - they are operating in a system where duplication is instant and frictionless, cleanup is manual and time-consuming, and deletion feels risky and irreversible.

As a result, files grow silently over time: duplicate frames and unused components accumulate, version histories expand indefinitely, and performance degrades only after scale becomes a problem. There is no system signal connecting everyday actions to their impact on file performance, infrastructure cost, or environmental footprint.

Duplication is effortless (Cmd+D); cleanup is not

Deletion feels risky without clear recovery signals

Research Insights

Understanding how waste accumulates at scale

I combined self-audit, platform research, and external validation to understand how and why digital waste persists.

Waste is invisible until failure Designers only notice waste when files become slow or unusable—by then, accumulation is deeply embedded in their habits and hard to undo.

Defaults drive behavior Quick duplication is effortless—a single shortcut—while cleanup requires deliberate intention and dedicated time most designers never actually set aside.

Fear prevents deletion Uncertainty around losing the right version leads to "just in case" saving habits that compound silently across every file over time.

Scale amplifies small actions across millions of files

Continuous background syncing compounds storage and energy use

Defining the Opportunity

Making system impact visible at the moment of action

The core issue is not awareness; it is timing. The system provides no feedback when duplication happens.

How might we make digital waste visible at the moment it is created, so designers can act early instead of reacting to performance breakdown?

Surface invisible impact Make file health and duplication visible in real time, so designers understand the cost of their habits before files become unmanageable.

Guide, do not force behavior Introduce lightweight, contextual nudges at the right moments—guiding behavior toward healthier file habits without interrupting or constraining creative work.

Automate where possible Where habit change alone isn't enough, automate cleanup in the background so files stay healthy without requiring deliberate manual effort from the designer.

Design principles: surface impact, nudge gently, automate responsibly

Designed to scale across teams, permissions, and workflows

Design Exploration

Designing a system, not a feature

Instead of a single tool, I designed Canvas Health Mode - a system of interventions across the workflow.

Moment of creation Duplication triggers immediate visual feedback so designers see file weight increase in real time and can make an informed choice before continuing.

Moment of accumulation The system tracks patterns of redundancy across sessions and surfaces accumulated impact before it compounds into a noticeable performance problem.

Moment of intervention When thresholds are reached, the system surfaces clear suggestions or automated cleanup options—both remain fully reversible so nothing is lost without consent.

Canvas Health Mode: overlays, nudges, and adaptive tools

Alternatives: manual tools ignored, hard limits disruptive, automation invisible

Test & Iterate

Validating clarity and behavior change potential

I tested early concepts to evaluate whether users understood duplication signals, how they responded to system nudges, and whether interventions felt helpful or disruptive.

Key learnings: visibility increased awareness immediately, users preferred suggestions over forced actions, and timing of interventions was critical to avoid frustration.

Testing focus: comprehension and perceived disruption

Iteration: adaptive system vs static default

Final Product

A system that integrates into existing workflows

Canvas Health Mode introduces three key touchpoints:

Canvas Health Visualization Real-time overlays surface duplication and file health directly on the canvas, giving designers immediate visibility into the cost of their current file structure.

Smart Archival Nudges Contextual prompts suggest archiving older versions at the right moment, making the storage and performance impact of taking action immediately clear.

Adaptive System Tools Embedded indicators and cleanup actions activate when thresholds are reached and remain fully reversible, so designers stay in control of every change.

System logic: versions > 90 days flagged; inactive files enter eco mode

Feedback loops: storage and CO2 savings over time

Impact

Aligning user behavior with system performance

User: faster, more responsive files; reduced manual cleanup effort; clear signals before performance issues occur.

Business (Figma): reduced storage and compute costs; improved performance at scale; more efficient infrastructure usage. Sustainability: less unnecessary data storage and transfer; reduced energy consumption across systems; waste prevented before accumulation.

Success metrics: duplicate frames %, avg file size reduction, storage/compute costs

Projected impact: preventable waste and CO2/year reduction at scale

Reflection

Designing behavior change through systems

This project pushed me to work within tight time constraints while designing a complex system - and the biggest lesson was the tension between output volume and narrative clarity.

I produced a large amount of work, but learned that framing and storytelling are as critical as the design itself. Connecting decisions across a system and condensing complex thinking into something immediately legible is a skill I'm actively developing.

I used Figma, Claude, and ChatGPT throughout the process for research support, outlining, and formatting. The experience also refined how I think about AI in a design workflow - useful for acceleration, but requiring judgment about when it helps versus adds overhead.

Reflection: narrative clarity is as important as system output

AI support: acceleration with careful judgment