Spot the Outlier: Why Averages Hide Where Revenue Is Lost.
How outlier analysis reveals the channels, regions and segments that quietly decide growth.
Read the analysis →Budgets are usually set one channel at a time. Investors learned long ago that the mix matters more than any single holding. The same logic applies to marketing.
Most budgets sit below the frontier: the same risk could earn more.
Most marketing budgets are set channel by channel. Each platform gets a target, each team defends its line, and the total is whatever the lines add up to. An investor would recognise the problem at once: nobody is managing the portfolio.
A fund manager does not ask whether one stock is good. The question is what the whole portfolio is expected to return, how much that return can swing, and how the holdings move together. The same three questions apply to a media budget.
Expected return is the familiar part: what a channel earns per unit of spend, at the level of spend you plan to run. Risk is less familiar. Some channels deliver a steady return every week. Others swing with auctions, algorithms and seasons. Two mixes with the same average return are not equal if one of them misses the quarter one time in three.
Correlation is the part most plans ignore. Channels do not work alone. Creator content raises branded search. Retargeting depends on the traffic that prospecting buys. Cutting one line can quietly remove the return of another.
The question is never whether a channel is good. It is what the channel adds to the mix you already hold.
Plot every possible mix by its risk and its expected return, and a boundary appears: the set of mixes that earn the most for each level of risk. Most real budgets sit below that boundary. They carry the risk of an ambitious plan and earn the return of a cautious one.
Moving to the frontier does not require more budget. It requires moving the budget you have. The practical limit is saturation: every channel returns less as spend rises, so the best mix is the one where the next unit of budget returns the same wherever it is placed.
A portfolio view also changes how targets are set. A single forecast line hides how uncertain it is. Simulating many plausible futures gives a range, and lets a board choose a plan by how likely it is to be met, not by how good it looks on a slide.
Charts in this article are modelled examples and do not represent any client.