In this article
Problem framing and metrics
How do you decide which problem to solve?
Start from an outcome the business wants, list the customer needs that would move it, then pick solutions and test their riskiest assumptions. An opportunity solution tree is a visual map of the paths a team might take to a desired outcome, and Teresa Torres's version has four levels: an outcome at the root, the opportunity space of customer needs and pain points, the solution space, and assumption tests (Product Talk, December 2023). Say which branch you chose and why.
Why it is asked: it shows whether you pick problems or only solve assigned ones. Design is one of the business and support interview tracks, and where this round sits in the whole process is explained in the interview loop overview.
How do you balance user needs and business goals?
Put both on one map instead of trading one for the other. Torres says the opportunity solution tree helps resolve the tension between business and customer needs by connecting a business outcome to customer opportunities (Product Talk, December 2023). In an answer, name a time the two pulled apart, the opportunity that served both, and what you cut.
Why it is asked: product designers are hired to hold both views. For story structure see the situation, task, action, result (STAR) order in the STAR method.
How do you define success for a design?
Choose metrics in categories, tied to a goal. The framework of happiness, engagement, adoption, retention and task success (HEART) gives five categories for user-centered metrics, with a goals, signals and metrics process for mapping product goals to measures (Rodden, Hutchinson and Fu, CHI 2010). Pick the one or two categories that match the goal, not all five.
Why it is asked: designers who cannot name a measure cannot show impact.
What does each HEART category measure?
Happiness measures attitudes such as satisfaction, visual appeal, likelihood to recommend and perceived ease of use, usually by survey. Engagement is the user's level of involvement. Adoption counts new users starting in a period, retention counts how many users from one period are still present later, and task success covers effectiveness and efficiency (Rodden et al., 2010).
Text description of the diagram
- Happiness measures attitudes such as satisfaction and perceived ease of use.
- Adoption counts new users starting in a period.
- Task success covers effectiveness and efficiency.
Why it is asked: it tests precise use of a common vocabulary.
Which metrics would you track for a new onboarding flow?
Map the goal to categories. If the goal is activation, adoption tracks how many new users start in a period, task success tracks whether they finish setup effectively and efficiently, and retention tracks whether a later share of them is still using the product (Rodden et al., 2010). Add a short survey for happiness. Say which metric you would watch first.
Why it is asked: it checks whether you connect a design change to a measurable outcome.
Prioritisation
How do you prioritise a long list of design ideas?
Score each idea on the same factors, then discuss the ranking. Reach, impact, confidence and effort (RICE) is a scoring method that multiplies Reach, Impact and Confidence and divides by Effort; Intercom's article describes impact on a scale of massive 3, high 2, medium 1, low 0.5 and minimal 0.25, confidence at 100, 80 or 50 percent, and effort in person-months across product, design and engineering (Intercom, January 2018).
Why it is asked: designers influence the roadmap, and the answer shows a repeatable method. See the worked numbers at the end of this page.
When would you ignore a RICE score?
When a dependency or a commitment makes a lower-scored item necessary first. The author of the RICE article says it should not be used as a hard and fast rule, and that dependencies or features needed to close certain sales can justify working on a lower-scored project first (Intercom, January 2018). Name the case and the reason openly.
Why it is asked: it tests judgement over formula.
How do you decide when to stop researching and start designing?
Stop when the next piece of research would not change the decision. Use the assumption tests of the opportunity solution tree: test the riskiest assumption first, with the cheapest experiment that could show it is wrong (Product Talk, December 2023). Ship a small version, measure it, and return to research if the signal is unclear.
Why it is asked: speed with discipline is the product-team trade-off.
Knowing it is half the job. Saying it is the other half.
Interview Clozo is an AI interview copilot for real interview rounds: it hears the question and puts a strong answer on your screen, and you say it in your own words. See the features or how it works.
Start Free Trial →Design systems and craft
What is a design system?
A design system is a complete set of standards intended to manage design at scale using reusable components and patterns, in the words of the Nielsen Norman Group. Its benefits include speed and scale, freeing designers for complex problems, a shared language and visual consistency (Nielsen Norman Group, April 2021, reviewed August 2026).
Why it is asked: most product teams run one, and designers are expected to extend it.
What belongs in a design system?
Nielsen Norman Group lists three parts: a style guide with implementation guidelines and design principles, a component library of reusable user interface (UI) elements each with a name, description, attributes, states and code snippets, and a pattern library of groupings, layouts and templates (Nielsen Norman Group, April 2021, reviewed August 2026). Say which part you have built or contributed to.
Why it is asked: it separates users of a system from contributors to one.
What is atomic design?
Atomic design is a methodology that builds an interface from five stages. Brad Frost names them atoms (basic elements such as labels, inputs and buttons), molecules (small groups of atoms working as a unit, such as a search form), organisms (relatively complex components such as a header), templates (page-level layouts) and pages (templates filled with real content) (Frost, Atomic Design, chapter 2).
Why it is asked: it is a common way to explain component hierarchy.
What are design tokens?
A design token is information with a human-readable name, at minimum a name and value pair, and it may carry extra properties such as a type or description, according to the Design Tokens Format Module (Design Tokens Community Group, 2025.10, Final Community Group Report, 28 October 2025). That version states that the specification is considered stable, the format's first stable release; it is a W3C Community Group report, not a W3C Recommendation, so describe it as a stable community specification rather than a formal web standard. Colour, spacing and type sizes are typical tokens.
Why it is asked: tokens connect design files to code.
How do components and variants work in Figma?
A component has a main component, which defines its properties, and instances, which are linked copies that inherit changes from the main component (Figma Help Center, read October 2026). Edit the main component once and every instance updates, which is why teams build buttons and inputs as components. Variants group a component's versions, such as sizes and states.
Why it is asked: it checks hands-on fluency with the team's main tool.
Collaboration and critique
How do you run a design critique?
A design critique is a group conversation that analyses a design and gives feedback on whether it meets its objectives. Nielsen Norman Group recommends a presenter, critiquers and a rotating facilitator, a clear scope, shared objectives agreed first, work shared in advance, a short session, and comments tied to goals instead of opinions (Nielsen Norman Group, October 2016).
Why it is asked: critique is the daily collaboration ritual of design teams.
How do you take feedback in a critique?
Restate the objectives, tell the design's story concisely, avoid over-explaining and do not take feedback personally. Nielsen Norman Group lists these as presenter practices and warns against rushing to solve problems in the moment (Nielsen Norman Group, October 2016). Write down every comment and respond to the themes afterwards.
Why it is asked: it tests ego and learning speed. A related prompt is covered in tell me about tough feedback you received.
How do you work with engineers on constraints and handoff?
Bring them in early and design in the system's own components. A component library documents each element's attributes, states and code snippets, which gives designers and engineers one shared description of what exists (Nielsen Norman Group, April 2021, reviewed August 2026). When a constraint blocks the ideal design, ask which part of the experience matters most and agree an alternative.
Why it is asked: shipped quality depends on this relationship.
How do you handle disagreement with a product manager?
Return to the shared objective and the evidence. Nielsen Norman Group advises agreeing the problem and the user needs before giving feedback, since otherwise comments become subjective (Nielsen Norman Group, October 2016). Propose a small test where you can, and agree who decides. The product management hub covers the other side of the table.
Why it is asked: designer and manager disagree often, and the process matters more than the winner.
Presenting your work
Walk me through a project in your portfolio.
Tell it as a decision story: the outcome the team wanted, the opportunity you chose, two options you weighed, the one you shipped and the result in a metric. Keep roughly five minutes and leave time for questions. Name your own contribution with "I" and the team's with "we", as MIT's Career Advising and Professional Development (CAPD) guide recommends for stories (MIT CAPD, read October 2026).
Why it is asked: it is the main evidence of how you think. Opening framing is in tell me about yourself.
How would you approach a product-sense exercise?
Clarify the goal and the user, list their needs, generate a few solutions, pick one with a reason, and say how you would measure it. This is the opportunity solution tree in miniature: outcome, opportunities, solutions and assumption tests (Product Talk, December 2023). Say your assumptions aloud and name the first test.
Why it is asked: the exercise shows structure under time pressure.
A worked prioritisation
Three made-up ideas show how RICE ranks work. The inputs are illustrative, the scales are Intercom's, and the scores were computed in assets/data/rice.py.1 Simplifying the sign-up form reaches 2,000 people a quarter at medium impact (1) with 80 percent confidence and 2 person-months of effort, which scores 800. Redesigning the settings page reaches 500 people at massive impact (3) with full confidence and 3 person-months, which scores 500. Adding dark mode reaches 8,000 people at low impact (0.5) with 50 percent confidence and 5 person-months, which scores 400.
| Idea | Reach per quarter | Impact | Confidence | Effort (person-months) | Score |
|---|---|---|---|---|---|
| Simplify the sign-up form | 2,000 | 1 | 80% | 2 | 800 |
| Redesign the settings page | 500 | 3 | 100% | 3 | 500 |
| Add dark mode | 8,000 | 0.5 | 50% | 5 | 400 |
Text description of the diagram
- The RICE impact scale is massive 3, high 2, medium 1, low 0.5 and minimal 0.25.
- Effort is counted in person-months across product, design and engineering.
Follow-up questions
These are the questions interviewers typically ask after a portfolio story. The answers reuse the sources above.
How did you know the design worked?
Name the goal, the metric category and the result. The HEART framework from Google Research gives five categories to choose from (happiness, engagement, adoption, retention and task success), so pick the one your goal maps to and state the before and after (Rodden, Hutchinson and Fu, CHI 2010). If you had no number, say so and describe the signal you did use.
What would you change if you did that project again?
Pick one real decision and give the reason. A good shape is what you decided, what you learned afterwards and what you would do differently. Nielsen Norman Group's critique guidance frames feedback around shared objectives, so tie the change to a goal the team had rather than to taste (Nielsen Norman Group, 2016).
Can you describe a time engineering could not build your design?
Use the situation, task, action, result structure (MIT CAPD, read October 2026): the constraint, what you shipped and the outcome. Reusing existing components is one practical route, since changing a main component updates every instance (Figma Help Center, read October 2026). To prepare further, see interview loops, the behavioral interviews hub, questions to ask the interviewer and the UX designer interview questions.
- Choose metrics by category and tie each to a goal.
- Score ideas on one shared scale, then override it with a stated reason.
- Know the parts of a design system and who owns each.
- Run critiques against shared objectives, not opinions.
- Present portfolio work as decisions and results.
Sources
- Measuring the User Experience on a Large Scale: User-Centered Metrics for Web Applications — Google Research (Rodden, Hutchinson and Fu), CHI 2010, 10 Apr 2010
- RICE: Simple prioritization for product managers — Intercom (Sean McBride), 5 Jan 2018
- Opportunity Solution Trees: Visualize Your Discovery to Stay Aligned and Drive Outcomes — Product Talk (Teresa Torres), 6 Dec 2023
- Design Systems 101 — Nielsen Norman Group (Therese Fessenden), last reviewed 27 August 2026, 11 Apr 2021
- Atomic Design, chapter 2: Atomic Design Methodology — Brad Frost (Atomic Design, copyright 2016), 1 Jan 2016
- Design Tokens Format Module 2025.10 (Final Community Group Report) — Design Tokens Community Group (Final Community Group Report, 28 October 2025), 28 Oct 2025
- Guide to components in Figma — Figma Help Center (no page date shown; date = read date), 24 Oct 2026
- Using the STAR method for your next behavioral interview — MIT Career Advising and Professional Development (date = read date), 24 Oct 2026
- Design Critiques: Encourage a Positive Culture to Improve Products — Nielsen Norman Group (Sarah Gibbons), 23 Oct 2016
