Reporting defaults to a return figure or blended revenue, while the number leadership actually manages by is new buyers, and this year even the platform's own optimization target made that same switch. A test result priced in revenue cannot walk into that conversation. Whether a read can be built in cost per new buyer instead of return on spend has already come up, and the honest answer is that it takes more work, which means real budget spent on an answer denominated in the wrong currency, a spend that then needs defending rather than a finding ready to present.
A read priced in the same unit leadership already uses lets a result walk straight into the room instead of needing translation first. Converting between what the platform optimizes for and what the board decides on stops being necessary. The test budget gets spent once, on an answer that counts the first time it is shown.
Some campaigns pull in more leads than others, and ranking them on lead volume alone funds the wrong ones, because the campaign that fills the funnel fastest is often not the one whose leads actually enroll. When a budget recommendation moves money on lead count, each campaign needs re-checking by hand to see whether the extra leads convert, a check that recurs across nearly every campaign. That reconciliation puts measurement right back where it was supposed to be replaced: the call comes down to what sales feels good about that week, not a number anyone can point to.
A read priced in new enrollments per campaign, not lead count, stops a campaign strong on volume and weak on enrollment from outranking one that does the opposite. One number now covers lead flow and enrollment flow instead of two that disagree at the point of decision. And the hand-check on each campaign stops being needed before trusting the recommendation on the table.
It means the number a channel or campaign is judged against is the count of new buyers or new enrollments it produced, not a return figure or a raw lead count that can move independently of actual enrollment. A campaign can look excellent on one and mediocre on the other, and the new-buyer number is the one that should decide the budget call.
By building the causal read directly against enrollment or new-buyer counts instead of revenue, using variation already present in existing spend and enrollment history. It takes more setup than a standard return-on-spend read because the outcome is sparser, but it produces a number denominated in the same currency the business actually manages by.
By reading efficiency and enrollment growth from the same underlying number instead of two separate ones that can move in opposite directions. A campaign that looks cheap per lead but weak on actual enrollments is not efficient in the sense that matters, and a shared read is what exposes that gap before it shows up in results.
When a campaign does not have enough leads or enough enrollment history to separate a real difference from noise, when the decision needed is how to split budget between audiences inside a single campaign, or when day-to-day bid management is what is actually required. In those cases the honest answer is a wider range, not a confident number.
A campaign that looks strong on one metric and weak on the other stops hiding behind whichever number gets reported that week. Budget starts moving toward what is actually producing new buyers, and the case for moving it no longer depends on translating between two numbers that were never meant to agree.