The initiative was declared a success at the board meeting.
The pricing project shipped. Revenue was up nine percent year over year. The slide had a green checkmark. The room nodded. Everyone moved on to the next item, and the next initiative got funded on the strength of the last one.
Here is the question nobody asked. Did the pricing project cause the nine percent, or did the nine percent happen and the pricing project happened to be in the room when it did?
Nobody could answer that. Not because the team was careless. Because the company had no way to separate the signal from the noise. The market grew. Two large customers expanded. A competitor exited a segment. A price increase went out. All of it landed in the same quarter, and all of it rolled up into one number that went up. The story attached itself to the most recent visible action, and the most recent visible action was the pricing project.
I wrote recently about how a stalled value creation plan gets misdiagnosed. This is the other half of that problem. Before a plan stalls, it quietly fills up with initiatives that were never actually measured, each one funded on the confidence of the one before it. By the time the growth flattens, the company cannot tell which of its bets ever worked, because it never knew in the first place.
Why false confidence compounds
A wrong number is a problem you can find. A number you cannot evaluate at all is worse, because it lets you believe whatever you already wanted to believe.
When an initiative is declared a success on a narrative instead of a number, the cost is not just that one bad data point. The cost is the next decision. The board now believes pricing optimization works at this company. So the next budget cycle funds more of it. The team that ran it gets credibility and a bigger mandate. The playbook gets written down and applied to the next portfolio company. A single unmeasured win becomes the foundation for a series of bets, none of which is any better grounded than the first.
This is how a value creation plan accumulates unmeasured bets. Year one funds an initiative on the thesis. Year two funds the next one on the apparent success of the first. By year three the plan is a stack of conclusions, and underneath the stack there is no evidence, just a sequence of stories that each got told confidently enough to fund the next one.
The opposite failure is just as expensive. An initiative that actually worked gets buried because the quarter it landed in was a bad one for other reasons. The market softened, a big customer churned, and the initiative that quietly added three points of margin got cancelled along with the things that did not work. You killed the one bet that was paying off because you could not see it through the noise around it.
Both failures come from the same root. The company cannot decompose its own results.
What attribution actually requires
Attribution is not a sophisticated capability. It is three unglamorous things done before the initiative starts, not after it ends. The reason most companies cannot attribute is that they try to do all three after the fact, when it is already impossible.
A baseline before you start
You cannot measure a change if you did not measure the starting point. This sounds obvious and it is the single most common thing teams skip.
The pricing initiative needs a baseline that is more than last year’s revenue. It needs the trend revenue was already on, the segments it was growing in, the rate of customer expansion that was happening anyway, and the price realization the company had before anyone touched it. Without that, a nine percent increase is meaningless, because you do not know what the number would have been if you had done nothing.
The baseline has to be captured before the work begins, because once the initiative is live you can never reconstruct the counterfactual cleanly. Memory is not a baseline. A number written down in the system before the project started is a baseline. The difference between those two is the difference between attribution and storytelling.
A single agreed definition of the metric
The second requirement is one definition of the metric that every function signs.
Take revenue retention. Sales counts it one way, including expansions and excluding a few accounts they consider one-offs. Finance counts it another way, on a strict cohort basis. Customer success has a third version that nets out seasonal churn. When the initiative review happens, the meeting becomes an argument about whose retention number is right instead of whether the initiative moved retention at all.
If the metric is not defined identically across the organization, attribution is impossible before you even start, because there is no fixed target to measure against. The number moves depending on who pulls it. This is the same gap I described in the dashboard nobody opens, where the reporting exists but nobody trusts it, because the definitions underneath it were never agreed and every screen quietly tells a slightly different version of the truth.
The fix is not tooling. It is a decision. Pick the definition, write it down, get finance, sales, and operations to sign it, and use that one definition for the baseline, the target, and the review. One number per metric that nobody relitigates.
The ability to decompose growth
The third requirement is the hardest and the one that separates real measurement from theater. You have to be able to break a result into its parts so you can isolate what the initiative actually moved.
Nine percent revenue growth is not a fact you can act on. It is a sum. Break it down and it might be four points of market growth that you would have captured regardless, three points from two large accounts expanding for reasons unrelated to the project, one point of price realization that the initiative can credibly claim, and one point of new logo growth from a different team’s work entirely.
The initiative moved one of those nine points. That is a real and useful finding. It tells you the pricing work is worth maybe a fraction of what the board thinks it is worth, and it tells you where the rest of the growth actually came from so you can decide whether to fund more of that instead.
Decomposition requires that the underlying data carries the dimensions you need. Revenue has to be sliced by segment, by cohort, by new versus existing, by price versus volume. If the company can only see revenue as a single monthly total, it can never decompose anything, and every initiative will get credit for the whole movement of the line it happens to sit near.
The test you can run in one meeting
You do not need a project to find out whether a company can attribute. You need one question, asked about a past success, and the patience to listen to the answer.
Pick an initiative the company declared a win in the last year. Ask the team to walk you through the evidence that it worked. Then watch which way the answer goes.
If the answer is a baseline number, a clear definition, and a decomposition that isolates the initiative’s contribution from everything else moving at the same time, the company can measure value creation. Fund the next bet with confidence.
If the answer is a story about how the team worked hard, the project shipped on time, and revenue went up that quarter, the company cannot tell success from noise. It is managing the value creation plan on conviction, and every initiative in the plan carries the same risk as the one you just asked about.
Most teams give you the story. They give it confidently, because they believe it, and they believe it because nothing in their reporting has ever forced them to test it. That confidence is the problem. It is exactly the false confidence that funds the next unmeasured bet.
Build the measurement into the plan, not around it
The fix is the same discipline that makes a value creation plan diagnosable in the first place. Decide the handful of metrics the plan actually depends on. Define each one once, with every function signed on. Capture the baseline before the initiative starts. Build in the dimensions you need to decompose the result later.
This is the practical work behind a one-page data value creation plan. When every initiative is tied to one of a small set of agreed numbers, with a baseline and a target set before the work begins, the review at the end is not a debate and it is not a story. It is a read. You can see which point of growth the initiative actually moved, and which points came from somewhere else.
None of this is a twelve-month transformation. Agreeing definitions for five to seven metrics and instrumenting them is a quarter of focused work. The expensive thing is not the measurement. The expensive thing is three years of bets funded on stories, a plan full of initiatives nobody can evaluate, and an exit where the buyer’s diligence team asks which value creation levers actually worked and the honest answer is that nobody ever knew.
The initiative that looked like it worked is the most dangerous line item in the plan. Not because it failed, but because nobody can tell you whether it succeeded, and the whole rest of the plan was built on the assumption that it did.