Here's the most expensive mistake in software: building the wrong thing, beautifully. A team can spend a year polishing a product nobody actually needs — because they never asked. The UX (user experience) design process is a disciplined way to avoid that: it puts real users at the centre from day one and keeps checking with them, so you find out you're wrong early and cheaply instead of late and catastrophically.

The heart of it is a single idea — user-centred, iterative design. Don't guess, then build, then ship and hope. Instead: understand people, sketch a solution, test it with those people, learn, and go round again. Each loop makes the design a little more right.

The loop

UX design isn't a straight line from idea to launch — it's a cycle you repeat. Step through one full turn:

The arrow back to the start is the whole point. The first loop is quick and rough; each turn adds detail and confidence. You stop looping when testing shows the design works well enough — not when the calendar runs out.

The core activities

• User research. Before designing anything, learn who your users are and what they need. You interview them, watch them work, run surveys, and study how they currently solve the problem. The goal is to replace your assumptions with evidence. The golden rule: you are not your user — what's obvious to you, who built it, may baffle everyone else.
• Personas. Research produces a pile of messy notes; a persona turns it into a memorable, fictional-but-realistic character who stands for a real group of users — a name, a photo, goals, frustrations and context ("Priya, 34, a busy nurse who checks the app between patients on a cracked phone"). Personas keep the whole team designing for a specific human rather than a vague "average user."
• Task analysis. Break down exactly what a user must do to reach a goal, step by step ("to send money: open app → log in → pick payee → enter amount → confirm"). This reveals where the journey is long, confusing, or full of chances to make a mistake — the places your design must smooth out.
• Scenarios & user stories. A short narrative of a persona using the product to achieve a goal ("As Priya, I want to log a patient's reading in under 20 seconds so I can move on"). These keep features tied to real needs, not shiny ideas.
• Iterate. Design, test with users, learn, refine, repeat. The insights from evaluation feed straight back into the next round of research and design.

Why iterate instead of plan it all up front?

You might think a careful team could just design the whole thing correctly the first time. But you cannot reliably predict how real people will behave — they surprise you every time. The cost of a change also rises steeply the later you find it: fixing a flaw in a paper sketch takes minutes; fixing it in shipped code takes weeks and may break other things. Iterating means you catch the flaws while they're still cheap to fix. This connects straight to Norman's principles: testing is how you discover which affordances, signifiers and feedback your users actually need.