Is your tech stack ready for 2026 and beyond? Download our new eBook and get a blueprint for the composable AI data stack.
Many AI decisioning tools in martech optimize based on correlation, which scales the status quo and can automate the wrong strategy. Learn how causal data reveals what works, turning AI from a prediction engine into an impact engine.
In the video series, From Data to Decisioning, we explore the concepts of causation and correlation in marketing, and why causality data is the missing ingredient your team needs to drive meaningful growth with agentic AI.
Watch episode 3 below, and scroll to the bottom for a full list of episodes.
When it comes to building AI decisioning tools, most martech companies today are getting it wrong.
Not because the AI is bad. It’s because the foundation is backwards.
AI doesn’t question strategy… it scales it. So if your system is trained on the wrong kind of data, it will automate the wrong strategy at scale
Today, I’ll talk through why most AI decisioning tools today are automating the best of primarily suboptimal strategies — and how there should be a new approach to the concept of data on which AI is built.
Most AI decisioning tools today learn from historical patterns. They score propensity, find lookalike segments, and recommend ‘next best actions’ based on what correlated with success before.
But it’s all based on correlation — it’s pattern matching. “customers who did X tended to buy, so do more X.”
In other words, it’s predicting based on the status quo.
So what? This is how marketing has been done for decades.
The catch is: Prediction is not the same as influence. Just because the AI can predict what might happen doesn’t mean it can provide a recommendation that will change outcomes.
The risk with this operating model is that the AI could automate correlation at scale.
Customers receive repetitive experiences, you don’t see significant uplift with your campaigns, and your marketing spend could be completely wasted.
Causal data flips the question from ‘what happened?’ to ‘what works?’
It records the results of an action or treatment and compares it to a control.
Now, AI can evaluate alternatives: ‘If we do A instead of B, what changes?’ Using AI powered by that encodes causal relationships appropriately also gives marketers more detailed explanations and audits on campaigns — they know what worked, in what context, and why.
So, the problem isn’t AI. It’s AI trained on correlation.
Causal data turns AI decisioning from a pattern engine into an impact engine—one that learns from every interaction and gets better over time to compound your marketing growth.
Watch more From Data to Decisioning episodes:
Most marketing data is correlational. Learn how causality data from randomized experiments powers smarter AI decisioning and faster growth.
Most marketing data is correlational. Learn how causality data from randomized experiments powers smarter AI decisioning and faster growth.
Tobi Konitzer explains why marketing has become good at predicting behavior (correlation) but not reliably changing it (causation), and why “causality data” is the missing ingredient for AI to make decisions that actually drive growth.
