Budget decisions feel rigorous because they use numbers, but the rigor can be an illusion produced by measuring the wrong thing precisely.
A common allocation method is to compare surfaces on an efficiency metric, such as cost per acquisition or return on ad spend, then move money toward whatever looks most efficient.
The method appears disciplined. It can also reward surfaces that report well over surfaces that contribute most, and those are not always the same set.
The cleanest report is not always the highest contribution
Surfaces close to the transaction usually report clean efficiency because the sale happens on or near them. Surfaces that create or shape demand can report poorly because their contribution appears later and elsewhere, as a change in another surface’s conversion rate.
Compare both on the same immediate efficiency metric and the capture surface will often win, not only because it performed but because it was positioned to receive the credit.
Follow that comparison without adjustment and demand creation is gradually defunded to feed demand capture. The system can look increasingly efficient until there is less demand left to capture and the previously efficient surfaces begin to weaken.
Apparent efficiency can therefore be a sign of harvesting rather than growing.
Reward the harvester and starve the farm, and the portfolio eventually loses the source of its yield.
Start with the job the surface is meant to perform
The correction is to stop asking one question of every surface and begin with the job each surface is meant to perform.
A surface assigned to Create should be judged on whether it brought new demand into being. A surface assigned to Shape should be judged on whether it improved preference, confidence or the quality of the downstream decision. A surface assigned to Capture or Convert should be held to the efficiency and conversion measures appropriate to those jobs.
The role sets the evidence standard.
Allocation then becomes a portfolio of commercial roles rather than one ranking of incomparable numbers.
Core and Edge need different evidence standards
Core and Edge discipline extends the same logic.
The Core is proven commerce that deserves relentless improvement on the metrics suited to its role. The Edge is emerging demand that cannot yet prove itself on mature Core economics. Judged on the Core’s standard, the Edge will usually lose before it has had enough time or evidence to become useful.
Held in one portfolio, the Core funds confidence and the Edge buys learning.
The Edge is not exempt from accountability. It is evaluated on the quality of the hypothesis, the size of the controlled commitment, incremental change and speed to learn. The Core remains accountable to reliability, contribution and compounding improvement.
Give the next dollar a defined commercial job
Giving the next dollar a job changes the character of the budget conversation.
Allocation debates often collapse into an argument about whose number is more trustworthy. Assigning roles first introduces a better question: what job does the business most need done next, and which surface is positioned to do it?
That question can be answered through commercial logic, readiness, economics, signal quality, likely incrementality and speed to learn. It does not make the evidence perfect. It makes the decision more coherent.
The next dollar does not only need to be spent efficiently. It needs a job, a role and an evidence standard that fits the role.
Give it those, and allocation can move away from whatever reports most cleanly and toward what is most likely to grow the business.