Custom UX design.

UX design (user experience design) is about designing how a product works for the people who use it. It goes beyond the interface: it covers the research that precedes it, the structure beneath it and the iterations that follow launch.

A full UX process typically includes discovery and user research, information architecture, user flows, low-fidelity wireframes, mid-fidelity mock-ups, a clickable prototype, visual design based on a design system, interaction and motion design, an accessibility check and usability testing with real users. It concludes with a hand-off to the developers: specified Figma components, design tokens and interaction notes.

We are a team of builders and designers under one roof. That means the UX designer isn't designing for an external party who has to build it later: the developer is at the table from the research stage. For scalable product families, we work from a design system so that every new screen is faster to build and more consistent.

The area where this makes the biggest difference is the grey zone between design and build: animations that look fantastic in Figma but cause performance problems in code, interactions that work in a prototype but not in a mobile browser, or a visual pattern that scales to two screens but not to twenty-two. By putting designer and developer in the same sprint, we prevent these frictions from surfacing only during the build phase.

Our clients are typically in the mid-market segment: scale-ups, SMEs with a digital product, and mid-sized corporate organisations with several internal tools. For some clients we deliver UX as a standalone service; for others it is built into a larger development project. The same discipline, a different scope.

A UX project rarely starts with design. It starts with a conversation about what needs to change and why. In the discovery phase we map out the context: who the users are, which jobs-to-be-done they need to complete, what the competitive landscape looks like, and which business goals sit behind the brief. We do this through internal stakeholder interviews (product owner, sales, support, sometimes a C-level sponsor) and external user interviews.

Based on that input, we build the information architecture: the sitemap, navigation, taxonomy and content structure. This is the phase where most rediscoveries happen. Almost every existing product has navigation that has grown organically and no longer matches the user's mental model.

Next come the user flows and wireframes: schema-free logic from entry point to conversion, then low-fidelity sketches to test patterns, then mid-fidelity mock-ups in which layout and hierarchy become visible. Only once the flow works do we apply visual design: typography, colour, components, micro-interactions and animation. Ideally this is built directly on a design system, so everything we make is component-based and reusable.

The clickable prototype in Figma is the first moment the design feels like a product. We use it for usability testing: typically a handful of moderated sessions with participants from the target group, or unmoderated sessions via Maze or UserTesting when the hypothesis is narrower. What we learn feeds into the next design iteration.

The final phase is hand-off to development: Figma pages with component names that match the codebase, design tokens for colour and typography, an accessibility checklist per screen, and a notes document for interaction detail. In our model this hand-off is almost ceremonial, as the developer has already been reading along and knows what is coming.

Since 28 June 2025, the European Accessibility Act has imposed obligations on commercial digital services aimed at consumers. For public bodies, the Dutch Temporary decision on digital accessibility has applied for some time. In practice, both come down to one technical minimum: WCAG 2.2 at level AA.

For our designs, that is the standard, even where the law does not require it. In practice, this means: sufficient contrast between text and background, keyboard navigation as a first-class requirement, screen reader compatibility, visible focus states, forms with semantic labels, and error messages that are accessible both visually and programmatically. We use axe DevTools, WAVE and Lighthouse for automated checks, and manually walk through the key flows with a screen reader.

Beyond WCAG, the wider compliance landscape also deserves attention. The GDPR and the associated ePrivacy directive call for privacy by design, which in a UX context largely means choosing to design cookie consent as a fair flow rather than a dark pattern. The AI Act requires transparency wherever AI functionality is visible to the end user; this concerns labels, explainability and opt-outs. For large platforms, the DSA and DMA also apply, with specific UX requirements around recommendations, opt-outs and advertising transparency.

For clients in more heavily regulated sectors, such as finance, healthcare and government, we combine UX work with enterprise software development that supports the audit standards they need.

Figma and FigJam are our standard for design and collaboration. Components, variants, auto layout, design tokens and prototyping all live in the same tool, and it is easy for client stakeholders to take a look without installing anything.

For research synthesis and information architecture, we switch between Notion, Miro and Whimsical, depending on what the client team already uses. We do not impose a tool stack on clients who already have one.

Usability testing is done via Maze (for unmoderated, quick, prototype-faithful tests) or UserTesting (for moderated sessions with selected panels). For sessions on an existing product, Hotjar and Fullstory are the first stops, offering heatmaps, recordings and funnel reports.

For analytics after launch, we connect to what is already in place: PostHog, Mixpanel or Amplitude for product analytics, GA4 or Plausible for web. We run accessibility checks via axe DevTools, WAVE and Lighthouse, supplemented by manual screen reader runs (VoiceOver on macOS, NVDA on Windows).

Where it adds value, we use AI tooling: Anthropic Claude for synthesising research transcripts and first-draft copywriting, Midjourney or Stable Diffusion for moodboards and visual exploration. Always transparent, always under human control and, where relevant to end users, labelled in the interface as the AI Act requires.