Most fail because they start with tactics instead of a specific business question, working backward only after the tactic has already been chosen. The second most common failure is running too many half-committed experiments at once, so nothing gets a fair test and nobody can say confidently what worked. A close third is no clear owner and no weekly review, so a plan quietly drifts for two months before anyone notices it's off track.
Kando's approach is to force the business question first — what specifically has to be true for revenue to move — before any channel or tactic gets discussed.