The quiet parts of a redesign

We redesigned Fieldnote last spring. The part people saw took three weeks. The part that made it possible took four months, and none of it is visible in the before-and-after.

What the four months were

Mostly deletion. Fieldnote had accumulated eleven ways to display a list of research notes, because eleven times someone had needed something slightly different and building a twelfth variant had always been cheaper that afternoon than reconciling the existing ones.

You cannot redesign eleven things. You can redesign two. So the work was: read all eleven, find that nine of them were the same thing with different padding, prove it to the two people who were sure theirs was special, and delete nine.

A redesign is mostly an argument about what things are the same. The drawing comes after.

The four questions

I ran every surface through the same four, in order, and it did most of the work:

1. Who is looking at this, and what did they just finish doing? Half our screens turned out to have no clear answer, which is its own finding.
2. What is the one thing they need from it? If two people gave two answers, the screen was two screens.
3. What is here because someone asked once, three years ago? Usually plenty. Almost all of it can go, and almost nobody notices.
4. What would this look like if we could not add anything? The most useful of the four. Constraint reveals what you actually believe is essential.

Why it looked easy

By the time we drew anything, there were two list layouts, one detail layout, and a set of decisions that had already survived every argument. The three weeks of visible work went quickly because all the hard parts had been settled while nothing was being drawn.

This is the thing I would tell anyone about to start one. If the visual phase is difficult — if you keep redrawing because the content will not sit still — you have not finished the quiet part. Go back. The drawing is not the problem.