Explainer Marketing, Sales, Customer Experience & Communications · Market Research and Consumer Behavior · Issue 2026-W46
Before expanding retail-media spending: are you measuring attributed sales or additional sales?
A new industry study includes a Middle East and North Africa sample in the retail-media conversation. For advertisers, its value lies in questioning what a campaign dashboard can establish. We propose a distinction between attribution and incrementality, followed by a short measurement agreement that should precede any decision to expand spending.
A new industry study includes a Middle East and North Africa sample in the retail-media conversation. For advertisers, its value lies in questioning what a campaign dashboard can establish. We propose a distinction between attribution and incrementality, followed by a short measurement agreement that should precede any decision to expand spending.
What the new release establishes
On 3 November 2026, IAB MENA published an overview of Attitudes to Retail Media 2026, describing the first regional sample of 37 participants with MENA responsibility. The European findings highlight measurement and reporting gaps. We subsequently accessed IAB Europe's report without a form and read its methodology and measurement sections. Regional and European analytical bases overlap, rather than constituting independent markets statistically. Accordingly, this article examines a measurement decision without applying survey percentages to Saudi advertisers. [1][3]
That is both a useful development and an important boundary. Including the region makes local practitioners' questions more visible, but does not turn every European result into a regional finding. A buyer's reported concern also differs from an experiment demonstrating an advertising investment's effect. The publisher is an advertising-industry association: its overview helps explain an industry discussion, while providing insufficient evidence on its own to establish a measurement product's effectiveness or one retail network's superiority.
Two questions that can look similar on a dashboard
We suggest opening the marketing meeting with two separate questions. Which sales do the reporting rules attribute to this campaign? Which sales would not have happened without it? Answering them may require different data and methods. We use attribution to describe assignment of credit within a report, while incrementality concerns the difference against what would have happened without the intervention.
This is an analytical distinction for organizing a decision. It does not imply that any available experiment can identify causality without error. Choosing a method still requires scrutiny of the comparison, available data, implementation and uncertainty. A campaign dashboard's polished presentation cannot resolve those questions by itself.
When a supplier reports a high return on advertising spending, ask about the attribution window, spending and costs included in the return's denominator, customers and channels included in the measurement scope, how returns are handled and what new-to-brand actually means. Does it mean no recorded purchase within a specified period, or something more extensive? We impose no universal definition on retail networks. Both parties need to disclose the definition actually used. If definitions differ between reports, the larger number cannot settle the comparison.
Attribution can remain useful for diagnosing execution even when it does not establish incrementality. It may reveal differences within the reporting window, without establishing that every attributed purchase would disappear if advertising stopped. Conversely, an incrementality exercise restricted to one category or period may not explain each daily change in creative performance. The business decision should determine method fit, rather than one favored tool answering every question.
A hypothetical example changes the meaning of return
Assume spending of ten thousand currency units and attributed sales of forty thousand. Attributed return is four times spending. Now assume, solely for comparison, that a suitable estimate places incremental sales at ten thousand, with a contribution margin before advertising of thirty percent. Incremental contribution becomes three thousand. Subtracting advertising produces a loss of seven thousand. Every number is invented; this is not a real campaign, platform or sector result. [A1]
The example does not show that retail media loses money, or that estimating the counterfactual is straightforward. It shows that an attractive attributed-revenue figure may leave the profitability of expansion unanswered. If the advertiser has another objective, such as learning about demand for a new product, state that objective and fund the exercise as learning. A loss cannot become a profit merely because a dashboard indicator improves.
Finance should examine the margin assumption as carefully as the measurement estimate. A contribution calculation that excludes a relevant variable cost answers a different question from one that includes it. For this illustrative exercise, the stated margin is an explicit assumption. In a real decision, the team should agree which costs belong in the calculation before seeing which version supports the preferred campaign.
Agree on measurement before approving an increase
We propose a short record whose meaning marketing, finance and the supplier all accept. It identifies the decision, definition of sales, measurement period, sources of cost and margin information, limits on data access and the person authorized to stop the exercise. Add a specific hypothesis: which result would support expansion, and which evidence would lead the team to defer it?
This record is not a legal contract. Its purpose is to prevent the question changing after the results arrive. If the objective was incremental contribution, success cannot quietly become attributed revenue because that measure looks stronger. Equally, a learning exercise should not be dismissed simply because it was never designed to deliver immediate profit.
Start with a limited campaign that can be assessed within available permissions and privacy requirements. Without a suitable comparison, call the findings descriptive and retain that limitation in the executive presentation. Do not construct a comparison group of customers with fundamentally different motivations and declare the difference causal. Deferral may be appropriate when the team cannot verify sales definitions or obtain the costs needed to assess incrementality. Compare the cost of waiting with the cost of expanding a poorly understood decision.
The discussion worth having after the report
IAB's cross-channel measurement decision guide, released for comment in October, starts from the business decision when weighing modeling, experiments and attribution. Its comment period ended on 4 November. We use that public description as professional context and make no claim to have read gated sections or a final approved version. [2] Applying the idea to retail media requires transparency between parties, beyond renaming a dashboard column.
At the next budget meeting, place the largest reported number beside the question it answers, then write down the question it leaves open. Can your team explain the difference without its measurement supplier? To choose a community and follow the conversation, visit ARPS communities; its public page describes ordinary membership as free.
References
- Attitudes to Retail Media 2026 overview
- Cross-Channel Measurement Guide Public Comment Period Closed
- IAB Europe Attitudes to Retail Media Report 2026
References open in a new tab.
Discuss this with practitioners
Create a free account to save articles and join the Specialists community discussion.
Discussion 0 published comments
No published comments yet. Be the first to share your experience.