Against the discovery phase

If discovery takes a quarter, it is not discovery. It is a way of billing for indecision, and it produces a deck rather than a decision.

We are not against research. We are against research that has no stopping rule. A study with a question attached — how do people actually find room 4.12, what does the label look like at arm's length, which of these two names survives a phone call — ends the day it answers the question. A study with a theme attached never ends, because a theme cannot be satisfied.

If a document exists to make the fee feel proportionate, delete the document and reduce the fee.

In practice the useful part of discovery is almost always three things: read what the client has already written about themselves, talk to the four people who will have to live with the result, and go and look at the thing in the place it will exist. That is a fortnight, and most of it is walking.

The rest — the personas, the moodboards, the competitive matrix — tends to be produced for the meeting rather than for the work. If a document exists to make the fee feel proportionate, delete the document and reduce the fee.