Paid

Paid advertising is the only channel that turns money into attention on demand, which makes it the fastest to test and the fastest to waste. The mistake is not spending; it is spending before you know what you are testing and before you can tell whether it worked.

Three preconditions. You need a page that already converts somebody — from ENT-106 — because paying to send strangers to a page nobody responds to buys you an expensive confirmation of that fact. You need to be able to tell where a customer came from, or the spend teaches you nothing. And you need a budget you can lose entirely without it mattering, because the first round is tuition rather than investment.

Then run it as an experiment with a written stopping rule, exactly as ENT-110 would have you do. Decide the total you will spend before you start, and the number that would tell you it is working — enquiries, not impressions or clicks. Change one thing at a time. Give it long enough to produce a real number rather than turning it off on day two, and stop at the figure you set rather than at the figure you feel. And count the cost properly: the platform spend plus the hours you spent managing it, because a channel that consumes six hours a week is not cheap at any spend.