For interfaces judged by how they hold up under real use
UI/UX Design
A polished interface that confuses real users has failed at the one job design actually has. This practice treats usability evidence as the standard, whether the work is building a design system from scratch, diagnosing why users struggle, or settling an interface before engineering starts.
How a usability evaluation actually runs
Heuristics, real user testing and wireframes revised against evidence.
90 secSample frame · video pending01Before building a prototype
Low fidelity or high fidelity is a real decision.
Prototyping fidelity gets chosen out of habit far too often, when the actual decision it needs to support should decide it. Four questions settle which fidelity fits.
Is the flow itself still in question, or just the visuals?
A flow that might change entirely does not need pixel-perfect visuals yet. Low fidelity tests the structure faster and cheaper.
Who is reviewing the prototype next?
Internal stakeholders can often work from wireframes. External users judging real usability usually need something closer to the finished look.
How much does visual polish affect the test itself?
Testing checkout flow logic rarely needs final visuals. Testing whether a brand feels trustworthy genuinely does.
How much time is there before a decision must be made?
Low fidelity iterates faster under real time pressure. High fidelity earns its cost when the decision can wait for more polish.
| Low fidelity | High fidelity | |
|---|---|---|
| Best when | The flow itself is still uncertain | Visual and brand perception matters to the test |
| Speed to test | Fast, cheap to change | Slower, more costly to revise |
| Feedback quality | Structural, about the flow | Includes visual and emotional reaction |
| Typical audience | Internal stakeholders, early testing | Real users, later-stage validation |
02Agreed before design starts
The ledger that keeps a design engagement honest
Design engagements go wrong when 'looks good' quietly replaces 'tested and usable' as the actual standard. This ledger fixes what you bring, what we design, what returns, and what stays yours.
- Your side
You bring
- The existing product and every screen already shipped
- Brand guidelines, if any exist beyond a logo file
- The team who will maintain the system after handover
- Our side
We build
- A design system built from an audit of what already exists
- Interaction patterns made consistent across the whole product
- Component documentation your team can extend without guessing
- Your side
You receive
- A working design system your team can apply immediately
- Documentation covering every pattern and its intended use
- A rollout plan for applying it across existing screens
Never leaves you
- Brand direction and every creative decision above the system
- The pace of rolling the system out across the product
- Ownership of the system once it is handed over
Where the line sitsA design system only works if the team actually uses it. We build it from what already exists, not a theoretical ideal nobody will adopt.
03From audit to a testable prototype
Four stages, each closing on evidence
Designs that confuse real users usually skipped a stage here. This system closes each one with something evaluated against a real heuristic or a real person.
Audit the current state first
Existing screens, flows or complaints are reviewed honestly before any new design work starts, so effort targets what is actually broken.
What exists, honestly assessedMap information architecture and flow
The structure of the product is settled before visual design begins, because a beautiful screen inside a confusing flow still confuses people.
Structure before surfaceEvaluate against usability heuristics
Designs are checked against known usability heuristics deliberately, catching problems a purely subjective review would likely miss.
A heuristic review, not just an opinionTest prototypes with real users
Prototypes are tested with people outside the design team, at a fidelity matched to the decision the test actually needs to inform.
Tested with real people, not the team
04What actually moves the number
Four drivers, priced from the scope
Design work is priced against the depth of evaluation and testing needed, because these four factors change scope by multiples.
Size of the existing product to audit
A handful of screens audits faster than a sprawling product with years of inconsistent design decisions accumulated inside it.
Depth of usability testing required
Testing with a broad, diverse set of real users costs more than a lighter internal review of the same flows.
Accessibility requirements
Designing for a wide range of ability levels from the start costs less than retrofitting accessibility onto a finished interface later.
Design system scope
A full design system covering every component costs more than targeted patterns for a handful of high-traffic flows.
05Evidence and proof
Designs that fixed what users were actually struggling with.
Design proof is a usability problem named correctly and a fix that measurably helped. Every record here waits for written client approval before production.
Explore case studiesA heuristic evaluation found the exact screen losing half of new signups.
Pillar Learning had rising signup abandonment and no idea why. A heuristic evaluation and user testing found one confusing form step, and a redesign of that step alone recovered most of the lost signups.
- Signup abandonment
- Cut by roughly half
- Screens requiring redesign
- 1 of 14
A design system replaced years of visually inconsistent screens.
Orra Systems had grown to dozens of screens with no shared component library. A design system built from an audit of existing patterns gave the team one consistent set of rules to build from.
- Inconsistent components audited
- Over 40
- Design system adoption
- Full within one quarter
The evaluation named the exact step users were dropping off at, which we had guessed wrong about for months.
Sample · pending client approval
06Read this first
When commissioned design is the wrong spend
Commissioning a full design system for a small internal tool few people ever see wastes budget on polish nobody needed. The exits are marked plainly. Where design is justified, output is delivered ready for engineering in Figma or an equivalent standard tool.
The interface is small and rarely used
A simple internal screen used by a handful of people rarely justifies a full design system or extensive usability testing.
- Product discovery
The underlying idea has not been validated yet
Designing a polished interface for an unvalidated idea risks polishing the wrong thing entirely.
- Hire UI/UX Designers
You need a designer on staff, not a project engagement
If ongoing design work is the real need, hiring a designer into your team fits better than a discrete engagement.
| Shape | Best when | You direct | We carry |
|---|---|---|---|
| Focused evaluation | One flow needs diagnosing | What happens with the findings | The evaluation and revised design |
| Design system build | The whole product needs consistency | Brand and priorities | System design and documentation |
| Embedded designer | Design work is genuinely ongoing | Product and creative direction | Named designer and continuity |
07Straight answers
Asked before a design engagement is scoped.
01Do you design, or do you also build the interface in code?
Both, depending on what you need. Many engagements deliver design ready for engineering, while others hand off directly into an application build under this practice.
02How do you evaluate an existing product's usability?
Through a structured heuristic evaluation against known usability principles, followed by testing with real users where the evaluation flags genuine uncertainty.
03Do you build design systems from scratch or from what already exists?
From an audit of what already exists wherever possible. A design system built from real patterns already in use tends to see far better team adoption.
04How do you handle accessibility in design work?
By building it into the interaction design from the start, covering contrast, keyboard navigation and screen-reader behaviour rather than retrofitting it after the interface is finished.
05What does UI/UX design cost?
Priced against the size of the product being audited, testing depth required and accessibility scope. A targeted flow evaluation costs less than a full design system.
06Do you test prototypes with real users, or just internally?
Both, matched to the fidelity and the decision at stake. Structural questions often only need internal review; real usability judgements need real users.
07What if the product idea has not been validated yet?
That is usually a discovery question first. Designing a polished interface before the underlying idea is validated risks polishing the wrong thing.
08Can you hand off design directly to an engineering build?
Yes, design and application development share this practice's delivery discipline, so a validated design typically moves straight into engineering without a separate handoff.