Probabilities drive decisions. A company that’s 80% certain there’s a viable market for a new product will act very differently from one that’s only 20% certain.
Still, probability can be hard to grasp. It’s vague — a bit like beauty. We may not define it precisely, but we recognize it when we encounter it. We’re certain the sun will rise tomorrow, but we’re far less sure the sky will be blue. And somewhere in between sits our faith in the weather forecast and the certainty of marketing outcomes.
Marketing lives in a zone of uncertainty because markets are made of people. Consumers, buying groups and enterprises exchange billions of signals every day. These interactions create feedback loops and introduce unknowns into every situation.
Despite marketing’s VUCA (volatile, uncertain, complex, ambiguous) nature, it’s still common for managers to cling to the mistaken belief that with enough data, the right strategy or a top-tier analytics team, we could finally be sure what will happen.
Why seeking certainty fails in marketing
Avoiding a certainty-seeking mindset is tough. Human minds don’t think naturally in probabilities — quite the opposite. We love definitive causes and effects, and our brains search relentlessly for them. As small children, we learned to draw simple conclusions. Uncomplicated outcomes, such as how the dog got out, likely result from the sum of a few knowable causes:
Jack left the door open while bringing in the groceries + The dog loves to be outside = The dog got out.
Problems arise when decision-makers misunderstand the relationship between outcomes and their causes in complex environments like marketing. A reductionist thought process, like the one that figures out how the dog got out, assumes all results can be traced to a chain of understandable causes. Marketing outcomes do have reasons, but they aren’t the clear answers our brains crave.
Although we may try to produce a particular marketing outcome, all outcomes in a VUCA world derive from multiple contributing factors, each of which has only a probability of happening. Reductionism leads to poor decisions in complex environments for the following reasons:
| Flaw | Description |
| Causes aren’t completely discoverable. | All marketing outcomes derive from multiple factors and some information will inevitably be missing, despite diligent investigation.
Hidden drivers, such as an unknown influencer, may unexpectedly change a marketing outcome. |
| Causes aren’t always linear. | Some causes can be readily linked to an outcome because we can see the connection, but other causes were set in place long ago or far away and are only now having an effect.
Advanced analytics and AI help to see more patterns but won’t be complete. Non-linearity, sometimes called the butterfly effect produces connections that are impossible to see in advance. |
| Causes have varying impact. | The multiple factors that contribute to marketing outcomes influence each other.
Some causal effects carry more weight than others and some factors will diminish or amplify others. |
Because of these reasons, trying to establish a definitive, repeatable chain of causes for marketing outcomes will always result in disappointment.
Dig deeper: Why marketing benefits when it provides forecasted guidance
The better way: Think like a statistician
To think like a statistician is to give up seeking certainty in predictions and learn to work more productively with probabilities. Probabilistic thinking equips leaders to better assess risks, weigh scenarios and make more informed decisions.
Every cause contributing to a marketing outcome has a degree of certainty. However, because markets consist of people, this degree of certainty will be substantially lower than the 90%–95% confidence common in university statistics classes.
According to a report cited in Noise: A Flaw in Human Judgment, an inspection of 708 studies in the cognitive and behavioral sciences, which look for patterns in human behavior, found that only 3% of studies produced correlations that were .50 or greater. Any correlation greater than .50 is considered strong.
Thinking like a statistician requires a new mental framework. Four mental practices will assist decision-makers in successfully applying this shift.
Make peace with not knowing
While more data, better analytics and improved processes will increase certainty, some things will never be known — that is OK. Of course, decision-makers should reduce the amount of uncertainty to the lowest reasonable degree.
Even with more time, money or the best technology, complex situations like marketing will never reach zero uncertainty — no matter the stakes. Chasing certainty to an unreasonable degree or blaming people for what is inevitable helps no one.
Broaden your sources of information
Marketing and sales outcomes are rarely simple. They derive from multiple simultaneous factors. Tools like marketing mix modeling or causal AI help identify the right combination of variables that best explain a result. The more diverse your data sources, the better your chances of finding the best fit.
Place bets
Betting puts a price on beliefs and helps avoid risky opinion-based decisions. Placing several small bets, rather than one big one, makes more sense when things are highly uncertain.
Clarify ambiguity
Decision-makers will want to seek out objective, verifiable data when it is available. Still, many marketing decisions, such as determining brand values or deciding whether to promote someone, require judgment.
In these cases, the precision of decision-making can be improved by using clarity tools such as scales and benchmarks. You will get better outcomes if the decision-making group agrees on definitions.
To make better decisions in marketing’s messy VUCA world, think like a statistician.
Dig deeper: Why causal AI works when other forecasting models fail
The post Why good marketers think in probabilities, not certainties appeared first on MarTech.