Skills assessment

UX Designer skills assessment

A portfolio is the most polished artefact a designer will ever produce, and the least representative of the job you are hiring them to do. The job is messy: ambiguous problems, half-finished research, engineering constraints, and stakeholders who want the button bigger. In 2026 AI tools can generate a plausible-looking interface in minutes, so visual polish — already a weak signal — is now close to zero. A CV or a highlight reel cannot show reasoning, research literacy, or the ability to make good calls under constraint.

A strong UX assessment tests that judgement directly. Rather than 'redesign our homepage', it hands every candidate the same deliberately flawed flow — an onboarding, a checkout — and asks them to critique it, prioritise the issues under a stated constraint, and say what they would research before shipping. H-Evaluate generates that critique per job description, calibrated to your product and seniority, so no candidate has seen it on a prep site — and scores reasoning against a shared rubric, not pixels.

What to assess

The competencies that predict performance in this role, mapped to the five hiring pillars.

Practical / sandbox

Flawed-flow critique in practice

Working a realistic broken flow — identifying usability problems, proposing a constrained fix, and drafting a research plan — with AI tools available and their usage visible, so reasoning is scored, not rendering.

Domain knowledge

Interaction & IA reasoning

Structuring flows and content so the common path is obvious and edge cases are survivable, at the depth the role requires — with visual craft weighted explicitly to the actual role, not by default.

Cognitive

Problem framing

Turning a vague brief such as 'onboarding is confusing' into a specific, testable statement of who is failing, where, and why — and prioritising the issues that most affect the outcome.

Situational judgement

Constraint & engineering judgement

How a candidate handles realistic pressure — a legacy codebase, a two-sprint budget, feasibility pushback that costs 30% of what made a design good — treating it as input rather than insult.

Behavioural

Research literacy & collaboration

Genuine willingness to let evidence overturn their own design, and the ability to explain and defend that reasoning to stakeholders and engineers who challenge it, without either folding or bristling.

Practical / sandbox

AI-tool fluency

Directing AI that drafts flows and copy, spotting where the output quietly fails users, and editing it against research evidence rather than shipping a plausible surface unchecked.

How to structure the assessment

  • 1Use a flawed-flow critique — the same deliberately broken onboarding or checkout for every candidate — rather than an open 'redesign this' take-home that rewards free time and rendering.
  • 2Seed the flow with obvious problems, subtle ones, and a red herring that looks wrong but is a reasonable trade-off, and score prioritisation under an explicit constraint.
  • 3Deprioritise pixels in the rubric; weight problem framing, IA reasoning, and research instincts, and calibrate visual craft to the actual role in advance.
  • 4Allow AI tools and observe their use — designers use them on the job, and how a candidate directs and overrides AI output is itself a signal.
  • 5Run it in a monitored, time-boxed environment so conditions stay comparable across candidates rather than as an unsupervised take-home.

Signals that predict success

  • +Frames the underlying problem before proposing a redesign
  • +Prioritises issues by likely impact and names the trade-off explicitly
  • +Says what research would change their recommendation, not just what they would build
  • +Directs and edits AI output against evidence instead of shipping a plausible surface

Red flags to watch for

  • Retreats to visuals and adjectives when asked why a decision was made
  • Cannot name the constraints that shaped past work — or never noticed them
  • Treats feasibility pushback as philistinism rather than input
  • Leans on portfolio polish or an AI-generated mockup with no reasoning behind it

Assessment vs. interview

Design interviews drift into taste conversations faster than any other discipline — outcomes swing on whether the reviewer would have designed it the same way. A structured assessment gives every candidate the same flawed flow and a shared rubric, so taste bias and brand-name bias lose their grip and the interview can go deep on the reasoning the work sample surfaced. A portfolio shows what a designer made; the assessment shows how they decide — and you are hiring the decisions.

Skills assessment

Configure this assessment by role and seniority

Watch the emphasis shift in real time as you change the role and level — no signup.

Related reading

Frequently asked questions

How do you assess a UX designer?

Give every candidate the same deliberately flawed flow — an onboarding or checkout with real usability problems — and ask them to critique it, prioritise the issues under a stated constraint, and propose improvements with their reasoning. Score the reasoning, prioritisation, and research plan, not the pixels. It mirrors the real job and is hard to fake with a polished portfolio or an AI-generated mockup.

What skills should a UX designer test cover?

Problem framing, research literacy, interaction and information-architecture reasoning, constraint handling, and collaboration with engineering. Visual craft matters, but for most product roles the design system carries the visual load, so weight it explicitly and usually lower — decided before interviews start, not after seeing a portfolio you like.

Why not just review a UX portfolio?

A portfolio is a claim, not evidence. It shows outcomes without the decisions, team output without individual contribution, and final states without the constraints that shaped them. It is a useful conversation starter, but a structured work sample on your own flawed flow gives comparable, first-hand signal — and neutralises brand-name bias, where famous-logo work benefits from teams the candidate did not build.

Should a UX assessment test AI fluency?

Yes, as one dimension. Designers now work alongside AI tools that draft flows, generate copy, and produce plausible screens. The differentiating skill is judgement: knowing what to ask for, spotting where the output quietly fails users, and editing it against research evidence. Let candidates use AI in the work sample and score how they direct and correct it, rather than banning tools the job includes.