A Project, Week by Week

People like to know what they're buying. So here is a six-week project, laid out in the order it really happens, with the weeks marked and the quiet parts included.

The first three weeks

Week one: listen and measure. We visit the site twice. The first time we only look. The second time we measure, photograph and ask questions of anyone who works there. The brief gets marked up and sent back (see how we question one). By Friday we agree what the project is.

Week two: the plan. No renders yet. A floor plan, drawn by hand and then cleaned up, with three versions. We show all three. One is the obvious answer, one is the cheap answer and one is the answer we'd choose. The client usually picks a mixture. That mixture is the plan.

Week three: materials and light. This is the week the sample board comes out. Materials are laid on a table in daylight, and we keep them there. Lighting is drawn at the same time, never later. A room lit well can survive modest materials, and the reverse is rarely true.

The last three weeks

Week four: detail. The drawings multiply. Joinery sections, junctions, the places where one material meets another. This is where a project becomes good or ordinary, and it's the least visible part of the job. A client seldom sees these drawings, but they feel the result every day.

Week five: pricing and people. We send the drawings to two or three builders and sit down with them. Not to haggle, but to ask where the hard parts are. A builder's early comment on a junction can save a month. We choose partly on price, partly on who asked the best question.

Week six: handover. The site is built by then, or nearly so. We walk through with the client, a pencil in one hand and a list of fixes in the other. There's always a list. A good project has a short one.

What goes wrong

Mostly three things, in no particular order:

• A delivery arrives late and the schedule bends.

• A wall opens and shows something nobody expected.

• The client has a new idea in week four.

The third is the most welcome, and the most dangerous. We keep a rule: new ideas are fine, but each one has to take something out. A project is a fixed bag. Put in a thing and you remove a thing.

The thing nobody sees

Quiet weeks matter. The week we spend sitting with a drawing and not changing it is often the week we find the mistake. A studio that never pauses ships its errors.

Six weeks is short. We've learned to protect the pauses inside them.