The Attribution Model Is Lying to Your Leadership Team: And everyone in the room already knows it

Your marketing team is presenting performance data that does not match commercial reality. This is not incompetence. It is what attribution models do when they are asked to do a job they were not designed for.

Attribution is a measurement convention. It assigns credit for revenue to the touchpoints that can be measured, in proportions that were agreed at a specific point in time, using logic that made sense when the business was simpler.

It is not a financial audit. It is a narrative about what might have caused a sale.

The distinction matters because leadership is making budget decisions on the basis of that narrative as though it were proof.

Where the fiction sits

Last-click attribution records the final step of a customer journey that may have taken six weeks and eleven touchpoints. The channel that gets the credit is the one that happened to be last. The channels that built the intent over the previous six weeks get nothing.

Multi-touch attribution distributes credit more generously, but according to a logic nobody interrogates. Was the 20 percent weighting on first touch correct when it was set? Has it been reviewed since the channel mix changed? Nobody knows, because the model runs quietly in the background producing numbers that look precise.

Branded paid search is the specific example worth naming. In a business with established brand equity and significant organic traffic, a branded search campaign is capturing revenue that was already coming. The customer typed the brand name because they already decided to buy. The campaign is present. The attribution model credits the campaign. The ROAS looks strong. The actual incremental contribution of that spend is close to zero.

The incrementality question nobody asks

The question that changes every budget conversation is this: would this revenue have occurred without this spend?

Incrementality testing exists to answer it. It is not widely applied, because applying it honestly redistributes budget in ways that some channels, and some agency relationships, do not survive.

So it does not get applied. The attribution model runs. The numbers look reasonable. The leadership team makes decisions on a picture of commercial performance that is internally consistent and commercially incomplete.

This connects directly to why reporting systems drift toward confidence rather than accuracy. The attribution model and the reporting framework reinforce each other. Both are calibrated to produce a story the room will accept.

What this looks like in a real business

A B2C ecommerce business spending £2.4m annually on paid media discovered through incrementality testing that 34 percent of its attributed revenue from branded search would have converted organically. That is £816,000 in annual spend maintaining campaigns that were capturing existing demand rather than creating new demand. The attribution model showed strong ROAS. The commercial reality was that a third of the budget was producing no incremental revenue.

The acquisition team was not at fault. They were optimising against the metrics they were given. The metrics were producing confidence. The commercial system was producing waste.

What this means in practice

If your marketing reporting produces confident numbers and your leadership team still feels uncertain about what is actually working, the gap between those two things is information.

It means the measurement system is producing confidence rather than accuracy. Those are different outputs and they require different responses.

The right response is not a better attribution model. It is a structural read of what the commercial system is actually producing, underneath the model that is describing it. That is what a Marketing MRI examines: the gap between what the numbers say and what the business actually experiences.