Sketch before you pixel

Every project I take on still starts on paper. Not because it is romantic, but because a pencil is the fastest way I know to be wrong cheaply.

Paper keeps decisions loose

On screen, a rectangle with a shadow already looks like a decision. On paper it is just a box, and a box is easy to cross out. In the first week of a project I want ten bad ideas on the table, not one polished one that nobody feels allowed to question.

What I sketch first

• The one screen a person will use every day
• The moment something goes wrong, and how we recover
• The empty state, before there is any data at all

Those three tell me more about a product than any homepage ever will.

A page of messy boxes is the cheapest prototype you will ever make.

When to move to the screen

Once the team stops arguing about the structure and starts arguing about the details, it is time to open the design tool. That shift usually takes two or three sessions, and it saves weeks once we start building.