Paid
The $500 ad test that saved a $50k launch
Before you spend the launch budget, spend a fraction of it finding out which message people actually click.
Test the message, not the design
Run five headlines against the same image with $100 each. Ignore the click-through rate that flatters you and look at cost per signup. One of them will win by a mile. It is rarely the one the founder wrote.
Kill early, scale late
Anything that has not beaten the control after 48 hours gets switched off. The winner gets the budget, then a second round of variations. Boring, repeatable, and it works.
What it saved
On a recent D2C launch the winning message was the fourth of five. Had we launched with the founder’s favourite, the cost per order would have been roughly double. The whole test cost less than one day of the planned spend.