Point AI at the Wrong Question and It Will Confidently Lie
A model will answer whatever you ask with total confidence. The trick is asking a causal question, not a correlation one.
By Joris van Huët, Founder & CEOUpdated 2 min read
Confidence Is Not Correctness
An analytics model, however advanced, answers the exact question it was pointed at, so if you ask which ads correlate with sales you get a confident answer that is not the same as which ads caused sales. The confidence is free. The correctness is not.
Plenty of tools now wrap attribution in the language of intelligence and automation. The label is not the problem. The question underneath is. "Which touchpoints appear before conversions" is a correlation question, and a model will answer it beautifully and wrongly if what you actually needed was cause.
Why the Framing Decides the Answer
Ask a correlation engine which channels to scale and it will point you at the channels that show up near conversions, which are often the ones harvesting demand, not creating it. Optimize on that and you defund the top of your funnel. We walk through this failure in correlation is not causation in marketing.
The intelligence of the tool does not save you. A precise answer to the wrong question is still the wrong answer, delivered faster.
Ask the Counterfactual
The honest question is counterfactual: what would have happened to revenue if this channel had not run? That is causal inference, and answering it well is a modeling discipline, not a dashboard feature.
No black box. No vibes.
Frequently asked questions
Can AI solve marketing attribution on its own?
An analytics model, however advanced, answers the exact question it was pointed at, so if you ask which ads correlate with sales you get a confident answer that is not the same as which ads caused sales.
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
- AnalyticsAnalytics is the systematic computational analysis of data. It reveals customer behavior and measures campaign performance.
- AttributionAttribution identifies user actions that contribute to a desired outcome and assigns value to each. It reveals which marketing touchpoints drive conversions.
- Causal InferenceCausal inference is the set of methods for estimating how much an outcome changes because of a cause, rather than alongside it. It works by comparing what happened with an estimate of what would have happened without the cause.
- CausationCausation is the relationship where a change in one variable directly causes a change in another.
- ConversionConversion is a specific, desired action a user takes in response to a marketing message, such as a purchase or a sign-up.
- CorrelationCorrelation is a statistical measure showing a relationship between variables; it does not imply causation.
- 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.
- TouchpointTouchpoint is any interaction a customer has with a brand throughout their journey. In marketing attribution, each touchpoint is a data signal to understand marketing impact.