Marketing budget leakage cannot be resolved by renegotiating agency fees, reducing headcount, or consolidating technology vendors. Those interventions affect the cost side of the equation. The leakage addressed here operates on the output side: spend that is allocated, deployed, and reported as productive, but that does not reach the demand it was intended to influence.
The mechanism is attribution. Standard attribution models assign credit for conversions to the channels and touchpoints that were present in the recorded journey. A channel present at the moment of conversion receives credit. A channel that built the condition for conversion upstream, without a trackable last-touch event, receives none. The credit follows what the measurement system can observe, whatever actually produced the outcome. When budget follows credit, it concentrates in channels whose contribution is visible and away from channels whose contribution is structural but untracked.
This produces a stable error. Direct-response channels accumulate credit because they operate at the bottom of the demand chain, where tracking is reliable and conversion events are dense. Brand-building activity, category-level presence, and demand creation upstream receive diminished credit because their effects are diffuse, delayed, and not captured by the event-based logic of standard attribution tools. The budget allocation adjusts accordingly. Over time, spend concentrates in channels that claim credit efficiently, and channels that create demand lose budget.
Attribution models assign credit to observable touchpoints. They do not measure the counterfactual: what would have happened without the intervention. Spend allocated on the basis of attributed credit, without a measure of incremental contribution, is by definition leaking into channels that may be harvesting demand they did not create.
The research on this is direct. Gordon, Moakler and Zettelmeyer (2023, Marketing Science, Northwestern University and NBER) analysed 663 large-scale randomised experiments at Facebook and found that non-experimental attribution methods failed to reliably estimate causal advertising effects, even with access to over 5,000 user-level features. The implication is that the measurement systems used to allocate budgets across digital channels cannot distinguish between channels that caused a conversion and channels that were present when one occurred.
The commercial consequence is quantifiable. Research published in the International Journal of Computer Applications (Doshi, 2025) found that traditional attribution models produce budget discrepancies of up to 30% through systematic over-attribution to direct-response channels. A 30% discrepancy on a £10 million marketing budget is £3 million allocated on the basis of a measurement artefact and without causal evidence.
Sales leakage in this context refers to the revenue that does not materialise because budget has been withdrawn from the upstream activity that creates demand, in favour of the downstream activity that measures well. The pipeline does not fill at the rate the spend would suggest, because the spend reaches the demand harvesting point and misses the demand creation point. The harvesting point looks productive in attribution reports and is structurally insufficient as a growth mechanism.
Demand harvesting channels capture intent that already exists. Demand creation channels build the conditions under which intent forms. Measuring only the former and allocating budget accordingly is a coherent response to an incomplete measurement framework. As an approach to revenue growth it is incoherent.
Resolving marketing budget leakage requires two separate instruments. The first is a demand-side scan that establishes where category-level demand exists, what share of it a given company currently captures, and what structural gap separates the two. The second is a causal measurement framework that isolates the incremental revenue contribution of specific interventions, separating what a channel caused from what it was present for. Neither is available in standard analytics platforms. Both are required before a budget allocation decision affecting either demand creation or demand capture can be made on defensible grounds.
Budget leakage is a measurement problem before it is a spending problem. The corrective action is to measure differently and to allocate on the basis of what the measurement reveals.
The leakage has a revenue figure.
Brand Demand Scan identifies where category-level demand exists and what proportion your current spend is failing to reach.
View Brand Demand Scan →Brand Demand Scan puts a revenue figure on the gap this note describes, from your own Search Console data. Leadership decides where the next euro in marketing goes against that figure, and the RoI² Framework verifies the profit each decision adds.