Most creative portfolio interview questions test the interviewer's taste, not the candidate's ability. A hiring manager who asks "talk me through this project" is inviting a rehearsed five-minute tour of a case study the candidate has already told forty times, and the person who tells it best gets the job. That's a selection process for presenters. Design hiring should select for people who can solve problems with colleagues, under constraint, when the brief changes halfway through. Portfolio assessment works when the conversation goes to the places the portfolio can't show you.
I came to this from the other side of the table. After twelve years in consumer PR, I spent my working life watching people pitch: journalists, clients, creative directors. The pitch is always polished. What tells you whether somebody's any good is what happens in the thirty seconds after you interrupt it. Design interviews have exactly the same weakness, and a lot of recruiters and hiring managers don't realise they're sitting inside it.
Why the portfolio walkthrough rewards the wrong people
A portfolio is an edit. Every candidate has chosen what to show, cut what went badly, and written a tidy narrative around work that was almost certainly messy at the time. The final screens hide the three rounds of feedback from a nervous account director, the developer who said the animation was impossible, the client who changed the logo colour on launch week.
So when you ask for a walkthrough, you're marking a piece of storytelling. Fine, storytelling matters, designers present to stakeholders constantly. But it's one skill, and it's the one most easily coached. Plenty of candidates have been through bootcamp critiques and mentor sessions that taught them precisely how to narrate a case study. The ones who haven't, often the stronger practitioners from agencies where nobody had time to write things up, look worse on paper and worse in the room.
If you want a fuller grounding in what sits on the page itself, the earlier pieces on creative portfolio assessment and what recruiters look for in creative portfolios cover that ground. This piece is about the conversation that follows.
Beyond Tick Boxes: Diversity Recruitment Strategies That Actually Transform UK Workplaces
Master the Virtual Hot Seat: 7 Video Interview Techniques Recruiters Don't Tell You
AI Governance Careers Surge in 2026: Why This Role Will Define the Next Decade
Ask about the work that isn't there
The most useful single move in a portfolio interview is to ask about what was cut. Which version did you present first, and why did it die? What did you drop to hit the deadline? Which piece of feedback did you disagree with, and what did you do about it?
These questions have no rehearsed answers because nobody builds a case study around their own abandoned work. A candidate has to reach back into memory, and what they reach for tells you plenty. Someone who can describe a rejected direction with some fondness and no defensiveness has probably been through real critique. Someone who says nothing was ever rejected is either lying or hasn't worked on anything with consequences.
Try a few of these, and notice I'm not offering a script, because a script is just another thing candidates learn to game:
- What was the first idea you were sure about, and when did you stop being sure?
- Who on the project had the least patience with your work, and what were they right about?
- If you had another week, what would you change, and what would you leave alone?
The third one is the quiet gem. Candidates who answer only with refinements to the visuals are telling you where their attention lives. Candidates who say they'd test the onboarding flow with five real users are telling you something different.
Probe the constraint, because that's where design happens
Design is the management of limits: budget, platform, accessibility requirements, a brand guideline written in 2019 that nobody's allowed to touch, a legal team with opinions. Strong portfolios hide those limits. Strong interviews pull them back out.
Ask what the brief actually said, then ask what the brief should have said. Ask what the candidate would have done with twice the budget, and then half. Ask who signed it off. In my own world, the best campaign people could always tell you exactly which of their ideas survived the client's risk appetite and which died in a legal review, and they told that story without bitterness. Designers who can do the same are showing you something a Behance page never will: they understand that the work lives inside an organisation.
Accessibility is a clean test here. A candidate who has genuinely worked to WCAG standards will talk about contrast ratios and keyboard navigation without being prompted, as ordinary parts of the job. A candidate who treats it as a compliance afterthought will go vague. You can read the W3C's guidance yourself in an afternoon, which is worth doing before you ask, so you can tell the difference.
Collaboration is a skill you can hear in how people say "we"
Listen to the pronouns. A designer who describes a project entirely in the first person is either a freelancer, a soloist, or someone who's quietly taking credit for a team. Neither is automatically a problem, but you should know which one you're hearing.
So ask directly: who else was in the room, and what did each of them need from you? Ask about the handoff to development, since it's where a surprising number of otherwise lovely work falls apart. Ask what the engineer pushed back on and whether they were right. A designer who says "the developer was difficult" with a sigh is giving you a warning. One who says the developer saved them from shipping something that broke on older Android devices is giving you a hire.
And ask for a time they changed their mind because of someone junior's input. It's an awkward question. That's the point.
The strongest objection: structured questions can flatten the weird ones
There's a real argument against all of this, and it comes from the creative directors I respect most. A tight interview structure, they say, rewards conformity. It favours the candidate who can produce a neat story about process and constraint, and it penalises the strange, instinctive talent whose best work came from a hunch they can't explain. Plenty of excellent designers are poor at articulating how they work. Make the interview a process audit and you'll hire project managers with Figma licences.
I'll concede the point, because part of it is true. A conversation framework can become its own rehearsed performance if you use it rigidly, and some candidates do think better with a pencil than a vocabulary.
But the objection assumes the alternative is neutral. It isn't. The unstructured portfolio chat isn't giving the instinctive talent a fair hearing, it's giving the confident talker one. Asking about a rejected direction or a developer's objection doesn't demand polished theory. It demands memory and honesty, and the quiet designers often have both in abundance. The remedy is to let the candidate show their thinking in whatever form suits them: a sketch on the table, a half-finished file, a screen share of the messy working document rather than the clean case study. Offer that, and the structure stops flattening anyone.
What to do on Monday
Pick two of the questions above and drop the generic walkthrough request. Brief your whole interview panel on what you're listening for, because three people asking three different questions and scoring on gut feel is where bias walks in. If you're hiring across a team, agree beforehand what a good answer to "what did you cut" sounds like, and write it down. The CIPD's guidance on inclusive recruitment is a decent starting point for building that consistency, and if you're sourcing candidates in the first place, OHUB's employer tools let you see people in their own words before the interview rather than only through a curated PDF.
The portfolio will keep getting prettier. Templates, AI image tools and polished case study builders have made it cheap to look senior. That's exactly why the interview matters more now, and why a question that can't be prepared for is worth ten that can. How many of the designers you've hired in the last two years could have told you, unprompted, what their best project got wrong?
