design-critique-case-studies
Structured design critique using the Liz Lerman Critical Response Process, plus deep-dives on Stripe, Linear, Notion, Airbnb, Figma, and Arc, and post-mortems of redesign failures like Snapchat 2018 and Sonos 2024. Use when running a critique session, giving design feedback, or learning from a shipped product.
- License 许可
- MIT
- 最近更新
- 2026-11-10
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- name
- design-critique-case-studies
- description
- Structured design critique using the Liz Lerman Critical Response Process, plus deep-dives on Stripe, Linear, Notion, Airbnb, Figma, and Arc, and post-mortems of redesign failures like Snapchat 2018 and Sonos 2024. Use when running a critique session, giving design feedback, or learning from a shipped product.
- allowed-tools
- Read, Grep, Glob
02 资源文件 3
- 参考文档references/critique-methodology.md42.7 KB
- 参考文档references/product-deep-dives.md61.6 KB
- 参考文档references/redesign-failure-analysis.md36.5 KB
03 SKILL.md 正文
Design Critique & Case Studies
Critique is the cheapest quality intervention in the lifecycle: a flaw caught in wireframes costs hours, the same flaw after launch costs months. It also raises the team's quality floor, builds shared vocabulary, and spreads senior judgment.
Critique Types
- Studio — scheduled, 4-8 people, 45-60 minutes; roles: presenter, facilitator, critics, note-taker. For major milestones.
- Desk — informal 1:1, frequent; still anchor feedback to a principle or user scenario.
- Async — Figma comments or a 3-5 minute recorded walkthrough; tag every comment [Question], [Concern], [Suggestion], or [Praise].
- Self — checklist before group critique: empty/loading/error states, hierarchy clear in 3 seconds, strongest affordance on the primary action, user language, WCAG AA contrast and target size.
Liz Lerman's Critical Response Process
Four steps, in strict order:
- Statements of Meaning — critics name what works and why (specific, not "looks nice"), so the presenter knows what to protect.
- Artist's Questions — the presenter asks about what they are unsure of, steering feedback to real concerns.
- Neutral Questions — critics ask non-leading questions to surface intent ("What was your rationale for...", not "Don't you think...").
- Opinion Time — opinions only with the presenter's permission, and grounded in a principle or evidence.
Critique by Fidelity (30/60/90)
| Stage | Focus | Out of scope |
|---|---|---|
| 30% concept | Right problem, information architecture, core flows, mental model | Color, type, pixels |
| 60% mid-fi | Interaction patterns, content hierarchy, edge cases (empty, error, overflow, truncation), systematic heuristic pass | Visual polish |
| 90% pre-handoff | Token consistency, micro-interactions, WCAG AA, responsive behavior, implementation ambiguity | Foundational architecture |
Anti-Patterns
- Design by committee — name one decision-maker before the session; everyone is heard, one person decides.
- HiPPO — the most senior person speaks last, or run an anonymous written first round.
- Vague feedback ("make it pop") — every comment must cite a heuristic, principle, or user scenario.
- Preference vs. evidence — separate them; when evidence is missing, flag the question for user testing.
- Bike-shedding — time-box by decision impact; the facilitator redirects to higher-impact topics.
Always ground feedback in principles, heuristics, or evidence. When principles conflict, name the tradeoff and recommend how to resolve it: user testing, business priority, or design principle hierarchy.
Routing
| Question | Read |
|---|---|
| Running a session: facilitator guide, formats, feedback frameworks (I Like/I Wish/What If, Rose-Thorn-Bud, Feldman, Six Thinking Hats), async protocols, the 10-dimension scoring rubric, receiving critique, critique types, full Lerman steps | references/critique-methodology.md |
| Why a product's design works: Stripe, Linear, Notion, Airbnb, Figma, Arc, Duolingo, Vercel, Apple Health, Discord, plus condensed analyses (Raycast, Supabase, Superhuman, Cal.com, Framer, Obsidian, Loom, and more) | references/product-deep-dives.md |
| Redesign post-mortems: Snapchat, Windows 8, Digg, Twitter/X, Google Plus, Sonos, Healthcare.gov, Reddit API, YouTube dislikes, Skype; redesign safety checklist | references/redesign-failure-analysis.md |
Cross-References
- Grounding comments in a heuristic, especially at 60%:
nng-ux-heuristics - Biases in the critique room (anchoring, confirmation, groupthink):
cognitive-psychology-ux - Settling disagreements with a usability test:
ux-research-methods - Design system component use at 90%:
component-patterns-code - Quantitative evidence for feedback:
ux-metrics-measurement - Dark patterns and manipulative content:
ux-ethics-content-strategy
04 同仓库更多 Skills
no-description
暂无简介
good-skill
Fixture skill that satisfies every rule.
broken-link
Links to a reference file that does not exist.
Bad_Name
Name violates the kebab-case rule and does not match the directory.
skill-authoring
Write, review, and validate portable agent skills in SKILL.md format. Use when creating a new skill, editing an existing one, or asked whether a skill follows the Agent Skills spec.
weigh-bags
Record scale readings for each bag — empty (tare), outbound, current, final — and reconcile them against the itemised inventory to find unaccounted weight. Use when the user gives a bag weight, says how heavy something is, weighs their luggage, or asks how close to the limit they are. Appends to bags.yaml and runs scripts/reconcile.py.