Again, I want to caution against drawing too much in the way of causal inference, but this scatter plot looks a lot like older scatter plots I have done.

FORBES: Want Growth? Getting More Flights Might Help

Causal inference requires appropriate techniques.

FORBES: Why Medicaid is a Humanitarian Catastrophe

Further, machine learning tools, which many companies claim they use to solve such causation inference, are fundamentally useless against this objective, as they are based on predicting correlations with no causal structure.

FORBES: The Dangers Of Cargo Cult Data Science