ChatGPT prompts for product design

Act as a senior product designer working with product and engineering. Create a design brief and interaction direction for [PRODUCT PROBLEM] for [TARGET USERS]. Use the evidence and constraints below; do not fill gaps with assumed user behavior.

Research and current workflow: [USER EVIDENCE]
Product and technical context: [PRODUCT CONTEXT]
Business goal and success signal: [SUCCESS MEASURE]
Constraints and non-negotiables: [CONSTRAINTS]

Ask up to 3 clarifying questions only if a required input is missing. Otherwise, produce:
1. A problem statement separating the user problem, business goal, and evidence. Name assumptions separately.
2. The primary job to be done, key user scenarios, and the moment that triggers each scenario.
3. A prioritized list of functional requirements and content requirements, each written as an observable behavior. Include states for empty, loading, error, permission-denied, and success where relevant.
4. A recommended end-to-end user flow in numbered steps, including decisions, recovery paths, and exit points.
5. A screen or component inventory with purpose, essential information, primary action, and accessibility considerations for each item.
6. A prototype research plan: target participants, five task-based prompts, success criteria, and what to learn before shipping.

Call out tradeoffs such as speed versus control, progressive disclosure versus visibility, and desktop versus mobile behavior when they apply. Do not prescribe visual style without evidence or brand context. Do not invent research findings, legal requirements, analytics results, or user quotes.

Before answering, self-check that each proposed requirement traces to evidence, a stated constraint, or a clearly labeled assumption, and that every critical flow has a failure or recovery state.

Create a usability-test plan for [PROTOTYPE OR FLOW] with [TARGET PARTICIPANTS]. Research questions: [QUESTIONS]. Product context: [CONTEXT]. Write a moderator guide with a neutral introduction, five realistic tasks, follow-up probes that do not lead participants, observations to capture, and success criteria for each task. Include a note-taking table format and a synthesis method that distinguishes repeated evidence from isolated feedback. Do not claim the prototype is validated before testing. Self-check that every task tests a stated research question and does not teach the interface.

Critique this product design for [USER GOAL]: [SCREEN DESCRIPTIONS OR PROTOTYPE NOTES]. Known user evidence: [EVIDENCE]. Constraints: [CONSTRAINTS]. Return findings by severity, covering information hierarchy, interaction clarity, error prevention, accessibility, responsive behavior, and consistency with the stated goal. For every issue, name the user scenario, why it fails, and a specific design change. Separate evidence-based findings from design hypotheses. Do not comment on visual taste alone. Self-check that no recommendation contradicts a supplied constraint.

Write a product-design handoff specification for [FEATURE]. Approved flow: [FLOW]. Screens and components: [DESIGN DETAILS]. Product rules: [BUSINESS RULES]. Produce a state-by-state specification covering data shown, validation, empty/loading/error/success states, permissions, analytics events, responsive rules, and accessibility behavior. List open questions separately from decided behavior. Use observable conditions, not visual shorthand such as “make it intuitive.” Self-check that each interaction has a trigger, result, and recovery path where failure is possible.