| name | edbx-bad-design-canvas |
| description | Use when a designer wants to stress-test a product idea for unintended negative consequences before launch, question whether a solution addresses root causes or applies band-aids, identify who is being harmed or excluded by a design, evaluate a product across cultural, social, environmental, and safety implications, or run a structured pre-launch bad design audit across 12 consequence categories. Apply the Bad Design Canvas to systematically uncover the unintended negative consequences a product or service might generate — from cultural appropriation and inequity to environmental impact and exploitation. Trigger this skill for any mention of unintended consequences, pre-launch ethical review, who does this harm, stress-testing a design, bad design audit, break my design, or when someone says "let's think about what could go wrong with this product." Also trigger for "bad design canvas", "consequence mapping", "negative impact audit", "12 consequence categories", or "what harm does this cause." |
| version | 1.0 |
| tags | ["ethical-design","audit"] |
Bad Design Canvas
Overview
Bad Design Canvas enables you to evaluate your product or service based on potential or unintended "bad" consequences. It is about uncovering the new problems our "solutions" might generate, questioning our own perception of the problems we seek to solve in the first place.
Products and services have a range of impacts when released into real-use contexts and society. However, it is difficult to identify potentially "bad" impacts or unintended consequences. The Bad Design Canvas provides a structured 12-category taxonomy that forces a team to look at their idea from every angle of potential harm.
The method is explicitly adversarial — it asks designers to argue against their own work. This is not pessimism; it is responsible design practice.
Method #09 — Bad Design Canvas. From Universal Methods of Ethical Design. Hashtags: #breakmydesign #evaluateoutcomes. See also: Ethics Canvas · Judgment Call.
Canvas credit: Matthew Manos, licensed under Creative Commons CC BY-ND license. Source: gum.co/baddesign
The core sequence is: intake the idea → fill all 12 consequence cells → summarize severity → redesign against the worst findings → produce a consequence statement.
Use This Skill When
- You want to stress-test a product idea for negative consequences before launch.
- You need to question whether a solution addresses root causes or applies band-aids.
- You want to identify who is being harmed or excluded by a design.
- You need to evaluate a product's cultural, social, environmental, and safety implications.
- A team is about to ship something and needs a structured "bad design" audit.
- Someone says "let's think about what could go wrong with this product."
- You want a comprehensive "nothing escapes" consequence review.
- You need to convert identified harms into concrete redesign actions.
Inputs
Provide as many of these as are available:
- The product, service, or design concept to evaluate (name + brief description)
- The target community or user population — who is this for?
- The problem the product claims to solve — what is it trying to fix?
- A specific consequence category to focus on (optional — the method covers all 12 regardless)
- Launch stage (optional) — pre-launch ideation, live product review, or post-incident retrospective
If the user provides only the product name, proceed — the method is designed to work from minimal input.
The Twelve Consequence Categories
The Bad Design Canvas covers 12 categories of potential negative consequence. Covering all 12 ensures nothing escapes. Most teams naturally focus on 2–3 and miss the rest.
| # | Category | Guiding Question |
|---|
| 1 | 🌍 Cultural Appropriation | In what way(s) is the idea borrowed from a culture you or your team are not representative of? |
| 2 | 🩹 Band-Aid | In what way(s) is the idea failing to recognize the root cause of the problem, instead serving as a temporary solution? |
| 3 | 🔒 Unfair Control | In what way(s) is the idea leading to unfair control over the user/customer? |
| 4 | ⚠️ Exploitation | In what way(s) does the new idea inappropriately expose or objectify the community it aims to serve? |
| 5 | 🔄 Inefficiency | In what way(s) is the idea creating new inefficiencies, unnecessary complexity, confusion, or delay? |
| 6 | 🌱 Environmental & Social Impact | In what way(s) is the idea using resources from finite sources or at risk of creating harsh conditions for workers? |
| 7 | 👥 Stakeholder Abandonment | In what way(s) does the idea fail to consider or address important stakeholders? |
| 8 | 🛡️ Decreased Safety | In what way(s) is the idea creating, or contributing to, unsafe conditions? |
| 9 | 🚫 Inappropriate | In what way(s) is the idea generally offensive or inappropriate? |
| 10 | 😐 Boring | In what way(s) is the idea the last, plain boring? |
| 11 | 🏠 Displacement | In what way(s) is the idea contributing to substitution or replacement? |
| 12 | ⚖️ Inequity | In what way(s) is the idea contributing to inequity? |
Full definitions, probing questions, and example responses for each category are in references/twelve-categories.md.
Workflow
This skill has one integrated workflow with five steps. Run them in sequence.
Step 1 — Intake
Collect from the user:
- What is the product or service idea?
- What problem does it claim to solve?
- Who is the intended community or user?
- Are there any known concerns already?
- What stage is it at? (pre-launch ideation / live product review / post-incident retrospective)
If the user provides only the product name and description, proceed. The canvas works from minimal input.
Step 2 — Complete the 12-Cell Canvas
For each of the 12 consequence categories, generate a specific response to the guiding question. Each response should be:
- Specific to the product provided — not generic observations
- Honest and unflinching — the point is to find the bad, not defend the design
- Graduated by severity:
🔴 Significant — likely, severe, and unaddressed
🟡 Possible — possible, moderate, or partially addressed
🟢 Low/None — unlikely, minor, or well-mitigated
For each cell, provide:
- The consequence category name and emoji
- The guiding question (repeated for clarity)
- A specific response grounded in the product being evaluated
- The severity tag
Coverage: All 12 cells must be answered. There is no "N/A" without justification. If a category genuinely does not apply, explain why and rate it 🟢 with a brief justification.
Tone variation for Cell 10 (Boring): This category is intentionally lighter — it is a creative quality check, not a moral concern. Treat it with a lighter, more playful tone. A boring product is not unethical; it is just uninspired.
See references/twelve-categories.md for probing questions that help generate specific responses for each category.
See references/severity-assessment-guide.md for guidance on rating severity.
Step 3 — Consequence Summary
After all 12 cells are complete:
- Count: How many cells are
🔴 Significant?
- Top 3: Identify and rank the three most serious consequences
- Systemic vs. Addressable: For each
🔴 finding, classify whether it is:
- Systemic — requires fundamental redesign of the product concept
- Addressable — can be mitigated through specific design changes without rethinking the core concept
- Consequence clusters: Identify categories that reinforce each other (e.g., Exploitation + Unfair Control + Inequity often appear together)
See references/systemic-vs-addressable.md for detailed classification guidance.
Step 4 — Redesign Recommendations
For each 🔴 Significant consequence:
- Name one concrete design change that would reduce or eliminate the harm
- Name one stakeholder who should be involved in addressing it
- Ask the hard question: "Does this consequence suggest the product should not be built at all?"
For consequences classified as Systemic, the redesign recommendation should address the root cause, not just the symptom. If a Band-Aid consequence is identified as systemic, the recommendation should question whether the product is solving the right problem.
Step 5 — Output
Compile the full session into a structured deliverable:
- Complete 12-Cell Bad Design Canvas — all cells answered with severity tags
- Top 3 Consequence Summary — ranked list of the most serious findings
- Redesign Recommendations — for every
🔴 finding
- Bad Design Statement — a single declarative paragraph:
"The most significant unintended consequence of [product] is [X], which harms [stakeholder Y] by [mechanism Z]. This requires [redesign action]."
- "Should we build this?" Verdict (optional but recommended):
- Ship with changes — 🔴 findings are addressable and redesign is feasible
- Rethink fundamentally — 🔴 findings are systemic and the core concept needs revision
- Do not build — the product causes harm that cannot be designed around
Facilitation Notes
Solo use: Work through all 12 cells systematically. Do not skip categories. The categories you want to skip are often the ones with the most important findings. Use assets/bad-design-canvas-template.md as a blank worksheet.
Team use: Give each participant 5 minutes to silently fill the canvas individually before sharing. Then compare canvases — the differences between individual responses are where the richest discussion happens. Group clustering works best on a shared surface (whiteboard, Miro).
Time budget:
- Quick scan: 20–30 minutes (Steps 2–3 only, reduced responses)
- Full session: 60–90 minutes (all steps, thorough canvas)
- Workshop: 2 hours (with group discussion and redesign brainstorming)
Stage-specific guidance:
- Pre-launch ideation: Use to shape the concept before committing to a direction
- Live product review: Use to audit an existing product and identify harms already occurring
- Post-incident retrospective: Use to understand what went wrong and what the canvas would have caught
Integration with Other EDBX Skills
- edbx-worrystorming generates free-form worries; Bad Design Canvas provides the 12-category taxonomy to organize and stress-test them systematically.
- edbx-inverted-behavior-model forecasts behavioral consequences; Bad Design Canvas maps the broader social, environmental, and cultural consequences.
- edbx-humane-design-guide maps sensitivity exploitation across 6 categories; Bad Design Canvas maps a broader 12-category consequence spectrum.
- edbx-fair-patterns remediates dark UI patterns; Bad Design Canvas remediates systemic design consequences beyond the interface.
- edbx-cider surfaces exclusionary assumptions; Bad Design Canvas maps how those assumptions lead to inequity, exploitation, and stakeholder abandonment.
- edbx-normative-design-scheme applies ethical theory lenses; Bad Design Canvas provides the concrete consequence taxonomy those lenses should evaluate.
- edbx-pledge-works formalizes commitments; Bad Design Canvas
🔴 findings are natural inputs for what to pledge against.
- edbx-anotherlens surfaces designer biases; Bad Design Canvas reveals the external consequences those biases produce.
Run Bad Design Canvas early in a project to shape concept direction, or late as a comprehensive pre-launch audit.
Guardrails
- Do not skip categories. All 12 cells must be answered. The categories you resist are often the most revealing.
- Do not defend the design. The canvas is an adversarial tool. Its job is to break the design, not protect it.
- Do not sanitize the output. If a finding feels uncomfortable to name, that is the signal. The hardest consequences to articulate are often the most important.
- Do not judge the designer. Bad consequences in a design are not accusations against its creators. The method is generative, not punitive.
- Treat "Boring" differently. It is a creative quality check, not a moral judgment. Keep the tone lighter for this cell.
- Address every
🔴 finding. A significant consequence without a redesign recommendation is an incomplete canvas.
- Use the Bad Design Statement as a decision artifact. It should be specific enough to guide the next design decision or stop the project.
Deliverable Quality Bar
A strong Bad Design Canvas session output:
- answers all 12 cells with product-specific responses (not generic observations)
- rates every cell with a justified severity tag
- identifies and ranks the top 3 most serious consequences
- classifies every
🔴 finding as systemic or addressable
- proposes a concrete redesign action for every
🔴 finding
- produces a Bad Design Statement that names the harm, the stakeholder, and the action
- includes a "should we build this?" verdict grounded in the evidence
Example Output (partial)
Product evaluated: A food delivery app that uses gig workers, surge pricing, and gamified delivery streaks
Stage: Pre-launch ideation
Cell 4: EXPLOITATION ⚠️ — 🔴 Significant
The app commercializes both gig workers and hungry users. Workers have no labor protections, benefits, or stability — they are the product being sold to restaurants. Surge pricing exploits users in food deserts who have no alternative. The gamified delivery streaks turn worker desperation into engagement metrics.
Cell 6: ENVIRONMENTAL & SOCIAL IMPACT 🌱 — 🔴 Significant
Gig worker conditions constitute harsh labor conditions: no healthcare, no minimum wage guarantees, algorithmic management with no human recourse. At scale, this creates a precarious workforce. Single-use packaging from every order adds environmental cost.
Cell 11: DISPLACEMENT 🏠 — 🔴 Significant
The app displaces local restaurant delivery infrastructure, replacing employed delivery staff with gig workers who earn less. It displaces cooking as a daily practice for time-pressed users, increasing dependency on delivered food.
Top 3 Consequences:
- Exploitation — workers and users are commodified
- Environmental & Social Impact — gig precarity at scale
- Displacement — local infrastructure and food culture eroded
Bad Design Statement:
"The most significant unintended consequence of this food delivery app is the systematic exploitation of gig workers, who are treated as disposable infrastructure rather than human stakeholders, which harms workers by removing labor protections, fair wages, and stability. This requires fundamental redesign of the worker relationship model — or the product should not be built in its current form."
Verdict: Rethink fundamentally — the core business model (gig labor + surge pricing) generates systemic exploitation that cannot be designed around without changing what the product is.