Before I designed websites I worked on buildings and urban plans. The two fields look unrelated until you watch how projects actually fail. They fail in the same order, for the same reasons.
Constraints arrive before inspiration
A site has a plot and a budget: a plot is the client’s existing content, brand and technical debt; the budget is the maintenance capacity after launch. Every project I have seen go wrong ignored one of those two facts early and paid for it late.
In planning, ignoring the site survey does not produce a braver building. It produces a delayed one. In web work, ignoring the content inventory produces a design that looks excellent in Figma and impossible in production.
Late changes cost more than early ones
The cost curve is brutally non-linear in both disciplines:
| Phase | Buildings | Digital products |
|---|---|---|
| Concept | A sketch and a conversation | A wireframe and a conversation |
| Detailed design | Redrawing a package | Rebuilding layouts and components |
| Construction | Variation order, delay, cost | Refactor, regression risk, delay |
| Occupied | Works on a live building | Works on a live product |
The lesson is not “never change”. It is “spend your uncertainty early”. Anything unresolved at handover gets paid for at the most expensive moment.
Systems beat objects
A building is not an object, it is a set of relationships: structure, services, circulation, light, maintenance access, the way people will actually use it in five years. A website is the same: content model, templates, build pipeline, hosting, editorial workflow, the way it will be maintained after the launch energy is gone.
Designers who only design objects produce things that photograph well. Designers who design systems produce things that survive.
What I carried over
- Survey before sketching. Read the existing content before proposing a structure.
- One source of truth. In planning it was the base drawing; on the web it is the content directory.
- Design for the maintainer. The person using this in year three matters more than the person approving it in week one.
I still think the best digital designers I know think like architects: they care about what happens after the launch photograph.
