Human-Centred Design
Design That Actually Works: How Human-Centred Practice Builds Better Products
Design has a credibility problem in enterprise. Too often it is treated as a step in a process — wireframes before development, visual design before handoff — rather than as a discipline that shapes strategy. The companies building the best products have figured out that design is not a deliverable. It is a way of working.
Research: The Work Before the Work
Most bad design decisions happen before a single wireframe is drawn. They happen when a team skips research and builds from assumptions.
Human-centred design starts with understanding: who are the people this product is for? What are they actually trying to accomplish — not what they say in a survey, but what you observe when you watch them work? What is the gap between how they describe their behaviour and how they actually behave?
The research methods that matter most in enterprise contexts: contextual inquiry (watching people work in their actual environment, not a usability lab); jobs-to-be-done interviews (understanding the underlying goal, not just the feature request); diary studies (capturing behaviour over time, not just in a single session); and quantitative analysis (behavioural data from analytics, not just self-reported preference).
Research is not a phase that ends before design begins. It is a continuous input that informs decisions at every stage.
Information Architecture: The Foundation Nobody Sees
Information architecture is the structural design of shared information environments. It answers: what categories does this product contain? How are they related? How do users navigate between them? What do we call things?
Poor IA is the most common root cause of usability problems in enterprise products. Users cannot find what they need not because the UI is ugly, but because the mental model embedded in the IA does not match how they think about their work.
Good IA is built from card sorting (how do users naturally group concepts?), tree testing (can users find things in the proposed structure without visual cues?), and progressive disclosure (revealing complexity only when the user needs it).
A well-designed IA compounds: every subsequent design decision becomes easier because the underlying structure is clear and agreed upon.
Design Systems: The Investment That Pays Compound Returns
A design system is a collection of reusable components, guided by clear standards, that can be assembled to build any number of applications. Done well, it is the highest-leverage design investment an organisation can make.
The components are the obvious part. The less obvious parts — the ones that determine whether a design system actually gets used — are the documentation, the contribution model, and the governance process. A component library with no documentation is a graveyard of well-intentioned work.
Design systems pay off when they reduce design and development time per feature, ensure consistency across products and teams, create a shared language between design and engineering, and allow a small design team to have large organisational reach.
The mistake most organisations make is starting with the visual layer — colours, typography, spacing — before establishing the interaction principles and component behaviours. Start with the hardest components (tables, forms, data visualisations) and let the simpler ones follow.
Prototype and Test Before You Build
The highest-value design activity is testing a prototype before a line of production code is written. A prototype that takes two days to build and surfaces a fundamental usability problem can save months of engineering time.
Modern prototyping tools (Figma, Framer) allow interactive prototypes that are indistinguishable from production for a usability testing session. Test with five to eight real users, not your internal team. Real users will immediately surface the assumptions you did not know you were making.
Usability testing is not a quality gate at the end of design. It is an iterative practice throughout. Test early concepts with low-fidelity wireframes. Test refined interactions with high-fidelity prototypes. Test the built product before launch. Test after launch with behavioural analytics and continued user research.
The goal is not perfection at launch — it is a system for continuous improvement grounded in real user behaviour.
