Critique a design~30 min

Critique a Design: Onboarding Flow Teardown

Early preview draft — not yet practitioner or accessibility reviewed.
Goal
Practice giving specific, actionable critique on a real multi-step flow, not just surface-level opinions.
Scenario
A B2C mobile app's 5-screen signup-to-first-value onboarding flow has a 40% drop-off between screen 2 and screen 3. You've been given screenshots of all 5 screens and the drop-off data.
Audience
A product team that needs to decide whether to redesign screen 3 or cut it entirely.
Constraints
You only have the screenshots and the drop-off number — no user research, no analytics beyond that one metric. Say explicitly where you're inferring versus where you have evidence.
Deliverable
A written critique (300-600 words) identifying likely causes of the drop-off, at least 2 concrete alternatives for screen 3, and what you'd want to measure to validate your theory.
Counts on your evidence profile as
A completed Skill Lab with its rubric and your deliverable
Accessibility requirement
Consider whether your proposed alternatives would work for a screen reader user with no visual context — call this out explicitly if you didn't evaluate it.
AI policy
AI use is allowed in a limited way for this lab — see details below. You can use AI to organize your notes or check your writing, but the actual critique judgments should be your own reasoning — this lab is specifically practicing that skill.
Rubric
  • Specificity: Points to actual elements/copy/flow decisions, not general impressions like "it feels cluttered."
  • Evidence vs. inference: Clearly separates what the drop-off data actually shows from what you're guessing based on the screens.
  • Actionability: Alternatives are concrete enough that someone could actually build one of them next.
Resources

Skills: Usability Testing, Interaction Design

Relevant to: Interaction Designer

Example responses

Different valid approaches with their own reasoning — not one "correct" answer.

Cut screen 3 entirely

Merge screen 3's single input into screen 2, reducing the flow to 4 screens. The drop-off likely isn't about screen 3's content — it's that users hit a 3rd screen before seeing any value and bail.

Why this works: When drop-off happens at a specific screen boundary rather than gradually across a flow, the *existence* of another step is often a bigger factor than what's on it.

Redesign screen 3, keep the flow length

Keep 5 screens, but add a one-line explanation of *why* this step matters before asking for the input, and pre-fill anything inferable from screen 1-2's answers.

Why this works: If the drop-off data doesn't clearly separate "too many steps" from "this specific step is confusing," it's equally reasonable to fix the step's own clarity before assuming length is the problem — flagging that ambiguity explicitly is itself part of a good critique.

Your work

Sign in to start this lab.