Explaining your attribution method to non-specialists
Four analogies that survive questioning and two that collapse under it. Describing a causal read without jargon and without misleading anyone.
Three platforms, three answers, one budget to defend. Your GA4 export already holds the answer, and the first finding is free and never leaves your browser. Get your first finding free
By Joris van Huët, Founder & CEOUpdated 2 min read
Two analogies survive questioning and two collapse, and using a collapsing one costs you the second half of the meeting.
The ones that hold
| Analogy | Why it holds |
|---|---|
| A medical trial with a control group | The counterfactual is the point, and everyone knows the shape |
| Weather forecasting with a probability | Uncertainty is intrinsic, not a defect |
The trial analogy carries the essential idea: you are trying to compare against what would have happened otherwise, and the strength of the claim depends on how that comparison was constructed. It also makes the ranking of designs intuitive, since everyone accepts a randomised trial ranks above an observational study.
The forecast analogy does the uncertainty work. Nobody thinks a seventy percent chance of rain is a broken forecast. It teaches that a range is a property of honest prediction rather than a hedge.
The ones that collapse
"It is like a smarter version of your dashboard." It is not smarter; it answers a different question, and the analogy sets up the expectation that the numbers should broadly agree. When they do not, the analogy has made you look wrong.
"It is AI." This ends the conversation without transferring anything, invites the objection that it is a black box, and is exactly the thing you would criticise in a vendor. The wider argument is in why vendors should explain methodology openly.
The four sentences that do the work
Where the range comes from: how much your spend actually varied in the window. What it will not do: score channels too small to separate from noise.
What to say about the design ranking
Observational is the honest label for a read built on variation already in your data. Say it, and say where it sits: below a randomised test, above a rules-based model.
The rule
Use an analogy that makes the limits visible. An analogy that makes the tool sound better than it is will be tested, and it will fail in front of the people you most needed to convince.
Related answers
Frequently asked questions
How do I explain causal attribution without jargon?
Four sentences: what goes in, what the estimate is compared against, where the range comes from, and what it refuses to measure. A medical trial analogy carries the counterfactual idea and a weather forecast analogy carries the uncertainty.Which analogies should I avoid?
Calling it a smarter dashboard, which sets up the expectation that the numbers should agree and makes you look wrong when they do not; and calling it AI, which invites the black-box objection and transfers nothing.Should I tell people the read is only observational?
Yes, and say where that sits: below a randomised test, above a rules-based model. People are far more receptive to a claim that ranks itself honestly than to one that implies it is definitive.
Go deeper: Causal attribution, explained.
Sixty-second versions of these ideas: Causality Engine on YouTube Shorts.
Have an audience of ecommerce operators? Earn 20% recurring commission by recommending the €99 read.
Keep reading
Terms in this article
- AttributionAttribution identifies user actions that contribute to a desired outcome and assigns value to each. It reveals which marketing touchpoints drive conversions.
- Control GroupA control group is the part of an experiment that does not get the treatment being tested, so its outcomes stand in for what the treated group would have had without it. In marketing it is the audience, or the set of regions, held back from a campaign.
- CounterfactualA counterfactual is what would have happened under conditions that did not occur, such as a customer's outcome had they not received the campaign they did receive. It is never observed, so every effect estimate has to stand in for it with a control group, comparable units or a forecast.
- DashboardA dashboard is a visual display of key information required to achieve specific objectives. It consolidates data onto a single screen for quick review.
- Multi-Touch AttributionMulti-Touch Attribution assigns credit to multiple marketing touchpoints across the customer journey. It provides a complete view of channel impact on conversions.
- Observational StudyObservational Study observes the effects of a treatment or intervention without controlling exposure. It does not use randomization, making it susceptible to confounding and selection bias.