Write the loss first

One sentence is missing from almost every test brief I have read: the one that would tell you the test had failed.

Write it before the test runs. One line: if this loses, what is true. Not a caveat, not a risk register. The plain sentence somebody will have to say out loud in three weeks.

Most hypotheses are not written that way. They are written so any outcome can be read as progress. The new page will improve the funnel. The creative will speak more directly to the audience. Read those again and notice no result contradicts them. If the number goes up, the hypothesis held. If it goes down, the market was soft, the sample was thin, the placement was wrong, and a great deal was learned. A test that cannot lose is not an experiment. It is a plan to keep going, dressed as enquiry.

The reason the losing sentence gets left out is not sloppiness. By the time a test launches, something has already been spent on it. A fortnight of the studio. A budget line. A meeting where somebody called this the priority. That money is gone either way, but the room does not read it that way. The room reads a losing result as an invoice for the spend, and it will find a way not to pay.

Writing the losing sentence in advance is the only thing I have found that gets in front of that. It is agreed before anybody is invested in the answer, by people who do not yet know which side of it they will be on.

The line is uncomfortable to write, which is the tell. If nobody in the room wants to write down what losing looks like, nobody in the room intends to lose.