A shortlist in 72 hours, contributing inside two weeks, and a 30-day guarantee if the fit is wrong.
72-hour shortlist · 30-day replacement guarantee · Month-to-month
Portfolio quality is a poor predictor of whether somebody will work out on your team, and it is almost the only thing most companies screen on.
A portfolio shows finished work, usually after unlimited revisions, often unconstrained by a real backend, and sometimes never built at all. It says nothing about the things that determine whether a designer succeeds in a delivery team: what they do when an engineer says the interaction cannot be built in the time available, whether their handoff can be implemented without a dozen clarifying questions, and whether they design the empty state and the error state or only the screen with four perfect rows of data.
The most common failure we are asked to fix is not ugly design. It is design that cannot be built as specified, so engineers improvise, the result diverges from the intent, and both sides conclude the other is difficult.
So we interview for the working relationship. It is harder to assess than visual craft, and it is what actually predicts the outcome.
A portfolio answers none of these questions, which is why we ask them directly.
Whether an engineer can build from their file without a meeting. Spacing, states and behaviour specified rather than implied.
Empty, loading, error, and the row with a ninety-character name. Where design and production most often diverge.
What they do when told something cannot be built in the time available. We introduce a constraint mid-conversation and watch.
Contrast, focus states, keyboard paths and labelling treated as part of the design rather than a later audit.
Extending existing components before introducing new ones, and sequencing change rather than proposing to replace everything.
A live conversation in English, every time. Design decisions have to be defended to engineers and to stakeholders, repeatedly.
The expensive design failures are invisible in a review and obvious in production.
The clearest example is unspecified states. A design shows a table with six rows. Production has none on the first day, four hundred after a month, and one row where somebody entered a ninety-character company name. None of those were drawn, so a developer improvises three times, and the interface acquires inconsistencies nobody chose. The same goes for loading, partial failure, and what a form does when the server rejects it after the user has spent four minutes filling it in.
Accessibility is the version of this with legal and commercial consequences. Contrast that fails at the ratios anybody will actually test against, form fields identified only by placeholder text, interactive elements that cannot be reached by keyboard, focus states removed because they were visually untidy. Retrofitting this across a shipped product is far more expensive than designing it in, and it is routinely discovered during a procurement review with a deadline attached.
Then there is the design system problem. A new designer arrives, finds the existing components inconsistent, and starts a redesign. It is genuinely well-intentioned and usually correct on the merits. The result is a product with two visual languages for eighteen months, because the redesign was never resourced through to completion, and the inconsistency they were fixing is now worse than when they started.
Designer (2–4 years). Produces screens and flows to a defined direction. Comfortable in Figma, works within an existing design system. Needs support on research and on defending decisions.
Senior Designer (4–7 years). Owns end-to-end flows including the states nobody asked for, runs lightweight research, maintains and extends the design system, and negotiates with engineering rather than handing over and hoping.
Lead / Principal (7+ years). Owns product design direction, sets the system and the standards, mentors, and can argue a feature down to something smaller and better in front of stakeholders who wanted the original.
Product design alongside a delivery team. Ongoing feature design with engineering in the room. The largest category and the one where collaboration screening matters most.
Redesign of an existing product. Usually triggered by accumulated inconsistency. Needs someone who will sequence it rather than start everywhere at once.
Design system work. Building or rationalising components and tokens so the team stops re-solving the same problems. Specialist, and undersupplied.
UX research. Genuinely a separate discipline. If you need interviews, usability testing and synthesis, say so, because many strong visual designers do not do this.
Design for a specific platform. Where conventions differ enough to matter. See iOS and Android developers, whose platforms carry their own expectations.
Week one. They use your product as a user would, and they read the existing components before proposing new ones. Expect a short, specific list of inconsistencies rather than a redesign proposal. Expect them to sit with an engineer and ask what is expensive to build, because that conversation shapes every decision afterwards.
Weeks two to four. One flow designed completely — including empty, loading, error and overflow states — and handed over in a form an engineer can build without a meeting. Expect at least one thing they simplified because the original was not worth what it cost to build.
The warning sign is a designer who proposes a full redesign in week two. They may well be right that the current interface is inconsistent. Proposing to replace it before understanding why it drifted, or who will maintain both versions during the transition, is how products end up with two design languages and no plan to converge them.
Hiring on visual portfolio alone. It selects for craft and presentation, both of which matter, and tells you nothing about handoff, constraint handling or edge cases. Add a working session to your process; it is the highest-value thirty minutes you will spend.
Not distinguishing research from interface design. Both are called UX. Someone who runs excellent user interviews may produce mediocre interfaces, and the reverse is at least as common. Decide which you need, and if it is both, accept that you are looking for a rarer person.
Excluding engineers from the interview. The people who will implement the work are the best judges of whether a handoff is workable. Have a designer walk an engineer through a real file and watch the exchange. It is far more informative than a portfolio presentation.
Treating accessibility as a later phase. Contrast, focus states, keyboard paths and semantic structure are design decisions. Deferred, they become a remediation project, usually discovered during procurement or a compliance review with someone else's deadline attached.
Rates depend on seniority, research depth, and engagement length. Tell us the role and we will give you a firm number.
Tell us the role and we will send a shortlist within 72 hours.
Tell Us the RoleTell us the role. Three to five profiles within 72 hours.