Nobody has to ask

Every organisation has an amount below which nobody has to ask.

Under it a manager decides and moves. Over it the same manager writes a case, sends it upstairs, and waits eleven days for a signature. Twenty thousand riyals, five thousand pounds, whatever the figure happens to be in that building, everybody who has been there a year knows it to the decimal.

That number, not the strategy document, decides how much a team actually tries.

Watch where the live testing sits. The new creative format, the small audience nobody has served yet, the landing page that argues instead of describing, the cheap trial of an offer two people have been disagreeing about for a month. Almost all of it runs under the line, funded from money somebody already holds, because the moment a test needs permission it needs a rationale, and a rationale before evidence is a confident guess with a slide behind it.

Above the line the money is not braver. It is less free. An approved project carries the reputation of everyone who signed it, which is why it can no longer change halfway: what gets built in month four is what was described in the paper, whatever the first three months taught. The large number buys certainty and gives up the right to be wrong. The small number buys the right to be wrong and gives up scale.

Which is how a business holds an ambitious strategy and a low ceiling at once, and keeps talking about a culture of experimentation while every actual experiment stays inside what one person can spend before lunch without telling anybody.

If you want a team to try more, the useful intervention is not a document about testing. It is arithmetic. Raise the amount nobody has to ask about.