Correlation does not equal causation.
If there is one takeaway from this blog, it should be that.
It is a mistake too many businesses continue to make when trying to predict what will happen in the future.
Let’s look at an example. Take the pharmaceutical industry. Using correlations might tell you people who get better after illness are usually those who take prescribed drugs.
That sounds logical enough. Yet just because something correlates doesn’t mean it has caused it.
For example, did the person actually respond well to drugs? Would they have recovered anyway because they are physically fit? Were the drugs wrongly prescribed?
Let’s use a slightly more extreme example to illustrate the point. Take sunscreen sales and shark attacks.
Sunscreen sales have been known to rise alongside shark attacks, but most people will appreciate one did not cause the other – and will know that, in fact, hot weather and more time spent at beaches did.
Making decisions based on correlation only works when things don’t change – but we live in a volatile world that is continually evolving.
Because things continually change, businesses need to realise why they are changing and act on that intelligence before making decisions.
And, crucially, correlation can only tell you the relationship between two variables. How many decisions do you make where there are only two influencing factors? Likely not many.
How do we learn the causal effect of something?
The answer could be in causal modelling.
Rather than failing as the world changes, causal models learn causal relationships that can withstand change and adapt to new or emerging situations.
They also work in partnership with humans, allowing for continuous knowledge sharing, and thus giving people the context necessary to solve real-world problems.
Crucially, they also allow for experimentation – and the ability to look into hypothetical worlds to test different scenarios.
When embedded in your business, causal models mean you can simply change one element of your process or workflow, there and then, and test the result.
They empower humans to make better decisions, and provide the intelligence needed to meet strategic goals.
At Faculty, we help our customers learn causal relationships to make the best informed decisions and achieve outlier performance.
Our software implements a scientific approach to show not just what is happening, but why.
Knowing why something is happening gives customers the highest degree of confidence in their forecasts and therefore their decisions. It is known as decision intelligence.
Because the future is not the same as the past – and correlation definitely does not mean that one thing caused the other.
Causal models are at the heart of our decision intelligence software, Frontier. Find out how it can help you understand not just what’s happening in your organisation, but why and what you can do about it.