Most media plans are assembled from the bottom up. Each channel team asks for what it can defend, the numbers are summed, and the total is called a strategy. The result is a portfolio nobody chose.
An operator builds the plan the other way. There is a business outcome, an amount of capital available to pursue it, and a set of instruments with different risk, latency and expected return.
Latency is a real cost
Search returns quickly and tells you little about demand you did not already have. Brand-building media returns slowly and changes what search is able to capture. Both are legitimate. Confusing them is not.
The discipline is to state the holding period in advance. If a channel is bought as a long-duration instrument, it should not be judged on a two-week report.
The allocation question
Every line in a plan should answer one question: what would have to be true for this to be the best available use of the next dollar? If nobody can answer it, the line is inherited, not chosen.