Most failed website redesigns don’t fail because of bad execution — they fail because the project was shaped by internal opinions rather than how customers actually use the site. Design thinking offers a simple, practical structure to avoid that trap.
1. Empathise: understand the real user first
Before touching layouts or colours, look at how people actually use your current site — analytics, heatmaps, support tickets, and direct customer feedback. What do visitors search for on your site? Where do they drop off? This stage is about evidence, not assumptions about what customers “probably” want.
2. Define: name the actual problem
Turn what you learned into a clear problem statement. “Our website looks outdated” is an opinion. “Mobile visitors abandon the contact form after the second field” is a problem you can actually design around. The sharper the problem definition, the more focused the redesign.
3. Ideate: generate options before committing
Resist jumping straight to a single design direction. Sketch multiple approaches to the defined problem, even rough ones, before narrowing down. This stage is where fresh ideas surface — the first solution isn’t always the best one, and comparing options early is far cheaper than comparing them after development.
4. Prototype: make it tangible before building
A clickable wireframe or mockup lets stakeholders and even real users react to the direction before a single line of code is written. Catching a confusing navigation structure at the prototype stage costs an afternoon; catching it after launch costs a rebuild.
5. Test: validate with real users, not the team
Internal stakeholders are the worst judges of usability — they already know where everything is. Test prototypes with a handful of actual or prospective customers, watch where they hesitate, and refine accordingly before full development begins.
Why this matters for the outcome
Design thinking doesn’t slow a project down when applied well — it prevents the much more expensive cycle of launching, discovering the redesign didn’t solve the real problem, and redesigning again a year later.