January 16, 2024 ↘︎

Are data management practices leaving you susceptible to cognitive bias?

Lax data management is leaving space for cognitive bias to creep in leading to misinterpretation of information and a lack of trust in using data to make business-critical decisions.  

That’s according to Lyndon Apthorpe, Senior Value Optimisation Advisor at Domo. Speaking to Digital Balance clients recently, Apthorpe said: “Marketing data is polluted.”

The source of this pollution, according to Apthorpe, is single-use reporting and single-use data extractions. By trying to make sense of information in one system or across systems, data is losing its power and value. The manual handling of data, data aging and latency leave a gulf for error. And the greatest concern is the scourge of cognitive bias.

“Cognitive biases live in the space where there’s opportunity for interference or opportunity for inferring what data is telling us rather than being structured and prescribed,” Apthorpe said.

By definition, cognitive bias is a systematic error in thinking that happens when processing and interpreting information. A myriad of biases fall under the heading of cognitive bias including confirmation bias when you favour information that aligns with your existing beliefs. Another, attentional bias, refers to the bias to pay attention to some things while ignoring others. Meanwhile, anchoring bias speaks to a tendency to rely on the first piece of information you learn.

According to data scientist Matthew Mayo, “Confirmation bias can lead to cherry-picking data and reinforcing pre-existing beliefs.”

There are plenty more biases where these ones came from. But whichever one you choose – and several could be at play at once – the end result is affecting the decisions and judgements you make. Decisions and judgements that should be better informed by the use of data.

With that in mind, cognitive bias could be losing your business money. A lot of money.

Data is meant to better inform the decisions we make. But if we’re not using it properly, it could have the opposite effect.

It’s little wonder, then, that a survey by Gartner found only 53 per cent of decisions being made in marketing are influenced by data.

So how can we get around these inbuilt biases?

According to the Harvard Business Review, we’re all susceptible to biases, especially when we’re fatigued, stressed, or multitasking. The advice to overcome it is to, “fight bias at the organisational level using choice architecture to modify the environment in which decisions are made.”

Domo’s Apthorpe reckons the first step is to create a data-informed culture. Establish foundations from the ground up by implementing a unified data architecture and a unified method of measurement. Robust governance is required as well as a culture of measurement and accuracy.

“Once we have these foundations in place, organisational data literacy goes up. And, ultimately, we start to build more confidence around our decision-making. And our data culture improves over time,” he says.

This data-informed culture needs to start at the top.

Apthorpe says, “You need to have a mandate coming down that anyone responsible for optimizing ROI there needs to be some form of agreed central source of truth that we can baseline our understanding against. without that, that just creates more space, more cognitive bias, more opportunity for numbers to be missed, misinterpreted, and more pollution to form.”

It doesn’t sound easy but reviewing your data handling practices to get around this issue is going to be infinitely faster than reprogramming decades of learned mental behaviour. And in the long run, it will save you a tonne of money.

DB logo
DB logo
DB logo