Three questions that kill bad campaigns.

Three questions that have killed more campaigns than they have approved. We ask them at the front of every engagement.

The first question is who is the campaign for. Not the persona. Not the audience description. The actual person. If the team cannot answer in a sentence, the campaign is a target acquisition exercise, not a marketing exercise. Usually the answer surfaces a deeper problem. The team knows who they think the buyer is. They do not know who is actually buying. A week studying who's actually buying fixes this faster than any creative brief can.

The second question is what does the campaign cause to happen. Not what does it say. Not what does it look like. What does it cause. A campaign that drives awareness causes nothing. A campaign that moves consideration causes nothing. The verbs that imply causation in marketing decks are mostly verbs that do not cause anything. The only verbs that matter are the ones that produce a behavior the company can observe. Attended an event. Submitted a form. Cancelled a competitor. Renewed early. Specific. Observable.

The third question is what would prove the campaign worked. A team that cannot name the proof condition before launch cannot evaluate it after. The proof condition has to be specific, time-bound, and visible to people outside the marketing team. “Pipeline grew” is not a proof condition because it is not visible enough to defend. “We added 47 named accounts in segment X by the end of Q3” is. The team that defines the proof condition before launch will run a different campaign than the team that does not.

Most campaigns survive the first two questions and die on the third. The death is useful. Either the campaign gets reshaped into something defensible or it gets cut, freeing the budget for work that has a thesis.

The questions sound rigorous. They are not. They are basic. Most campaigns get launched without a clean answer to any of them. The teams that ask them ship fewer campaigns. The campaigns they ship perform better.

Fewer launches.

Better launches.