The useful response to a demand for sales attribution is to establish what data and comparison design would support it. A campaign should commit to the strongest evidence it can produce and make the remaining uncertainty visible.
Confidence should come from the measurement design and the quality of the data.
Separate the claims
A delivery claim says the campaign reached people. A behavioural claim says a qualifying action was recorded. An attribution claim applies an agreed relationship between exposure and an outcome. A causal claim estimates what happened because of the campaign.
These claims require different support. Naming the intended conclusion early helps a team choose the right tracking, data access and comparison method. It also keeps the final report from stretching a weaker signal into a stronger promise.
Ask what the campaign can observe
Can exposure be linked to suitable outcomes? Does transaction data cover the relevant products, stores and period? Are the definitions consistent? Can a comparison or test design account for important alternative explanations?
If the answer is incomplete, define a smaller claim and a path to improve it. Qualified leads, redemptions or measured visits can still be useful evidence when clearly labelled. A visit remains distinct from a purchase.
Treat the baseline as a design decision
An increase after launch may also reflect seasonality, promotions, store changes or other marketing. A suitable baseline or comparison helps assess these explanations, but its quality needs review.
Describe the method, assumptions, observation window and uncertainty. If the design supports only an association, use that language. Avoid implying that a precise-looking number removes the need for methodological judgement.
Invest in better evidence over time
The first campaign can identify the data and operational work required for a stronger next test. Agree who owns transaction access, event quality, location validation and response handling.
At 2plus3, the evidence ladder is a way to structure that conversation. Select the level the brief can support today, then identify what would let the next campaign answer a more commercially meaningful question.
Before promising sales attribution
- Define the commercial outcome and unit of analysis.
- Confirm coverage and quality of the necessary data.
- Agree matching, time windows and eligibility rules.
- Choose a comparison design and document its limitations.
- State the conclusion that the evidence will and will not support.
Put it to work
What can genuinely be attributed to the campaign?
Measurement methods and product availability are agreed for each brief. Examples describe possible journeys; commercial outcomes require suitable evidence.