Data For Optimisation. Intuition For Direction. The Decision Framework

Predictive analytics fails at exactly the moments that matter most, and this episode explains what to use instead.

Every predictive model ever built is trained on historical data. It looks backwards to tell you what happens next. That works brilliantly right up until something happens that has never happened before, and then the model has nothing. Oliver walks through the last decade of black swan events that no algorithm saw coming, and the operators who felt them arriving anyway.

This is not an argument against data. Oliver runs PinPoint Media, a performance marketing agency that lives in the data every day. It is an argument against the uncritical worshipping of it.

He covers the three specific ways data misleads decision makers, the difference between optimisation decisions and direction decisions, and a five-step process for making the calls that models cannot make for you.

The framework is simple. Data for optimisation. Intuition for direction. Feedback loops for everything.

Key topics covered:

Why predictive analysis fails at inflection points

The three ways data actively misleads decision makers

Survivorship bias and why your data only shows you the survivors

The correlation and causation collapse that quietly wrecks strategy

Recency bias amplification in volatile markets

Why intuition is compressed data, not the absence of data