Step 5: Choose 1 - 2 Channels As Hypotheses

In a perfect world you want channels with the following values:

Targeting - High, Cost - Low, Input Time - Low, Output Time - Low, Control - High, Scale - High

You will rarely get a channel that fits those values perfectly so the question is how do you prioritize? This is where the work you put into Step 1 and Step 2 come in. Sort your channels using your constraints as a guide. If your biggest constraint is money, then look at channels that have a low cost as the first priority and come closest to the ideal values for the other attributes.