The Problem With Data

Data is not information, information is not knowledge, knowledge is not understanding, understanding is not wisdom and only with wisdom can we have insight.

It is natural to both fear the unknown and to have a strong desire to allay that fear. Lack of insight in both business and life can bring the best plans crashing down.

Big Data, Small Data, Thick Data, regression analyses, log analyses, control groups, p values, and significance – in this modern world of news, fake news, and data breaches, we are constantly presented with numbers that are designed to give information an instant gloss of credibility. People often try to burnish their claims by saying things like, “scientifically proven”, “you can’t argue with the numbers”, “if you can’t measure it then it isn’t true”, and so on.

But the simple fact is that it is not that simple. There is something quasi-mystical in numbers, which makes them both instantly trustworthy and the perfect tool to bamboozle people. The trick is to look behind the numbers and understand what is being measured and how. Furthermore, some things, especially anything to do with human beings, are not easy to measure with ‘conventional’ statistics. For instance: how do you measure the strength and intensity of a feeling or an intention? It is not like calculating the re-entry criteria for a spacecraft, for physics doesn’t have feelings.

Being at the top of the intellectual food chain can make us believe that we are best placed to see into this unknown, exploiting data to see what is really happening in the world around us. This belief is powerfully seductive. The solutions being sold to us prey not only on the fear of the unknown, but also on the seduction of knowing. The heady mixture of loss-aversion coupled with our natural availability heuristic, that is marketed to a worried audience, often causes them to grab at the passing offerings in the belief that the silver bullet is in there somewhere.

As ‘mere’ beings we easily fall prey to the idea that we are masters of the (our) universe; we use technology in the hope that it will allow us to control what we want to control. But the problems we face in exerting control don’t come from the technology. They come from us. We are blinded to our own fallibilities and mistake output for insight. We can get captured by the belief that the latest tech provides the truth, and is a legitimate insight into the future. The desire to believe this leads us to distort the data to fit our assumptions, and inevitably this also produces a distortion of reality. Famously, Nokia was warned by an ethnographer that the smartphone was coming. They insisted that this person was wrong because, they said, the information did not ‘fit the data’. We all know what happened to Nokia. Nokia who?

The analysis of data is a lagging indicator; it involves measuring the past, interpreting that past, and trying to predict the future, and that is a tough challenge. We conflate what we see with what we understand and how we think it should be. There is a reason that investment products carry the dire warnings about past success being no guarantee of future performance.

There is no doubt that we are much better at incorporating such ‘soft’ characteristics into measurement metrics. However, it is not as easy as a pure data-science approach. To build the effective tools (algorithms) requires a much more nuanced and wider understanding than that given by a blinkered approach which fails to incorporate Thick Data. This can only really be done by a multi-disciplinary team of people, whose skills might include behavioural science, ethnography, sociology, political science and psychology – to name just a few. Mathematicians, statisticians, data-analysts and programmers are certainly necessary, but it shouldn’t stop there.

It is often said that people are at the core of a business. Whether they are the customers or the staff, they are people, not machines. Knowing what people do is one thing, knowing why they do it is even more important. More importantly still is understanding why people are doing what they do. This requires much more information that merely what is being done and by whom. This is the context that only Thick Data brings.

Human beings do not act rationally and famously, they will lie like mad to researchers! Something shown in many studies about the issues in conducting studies on people. Additionally, they can rationalise their actions in a way that they are happy with. Knowing the value of how the social, societal and environmental factors influence the numbers is a step towards the sort of understanding that may have saved Nokia. For modern business leaders, who rely on data to inform their decisions, it is critical to understand the context of actions and the intentions that underpin the actions.

If you want to take the blinkered approach offered by an IT package and believe that it is a magic software tool will allow you to predict the future, then I would suggest that you are falling prey to unconscious bias. When that happens, you find things like the following flip-chart starting to seem credible. In fact, I’ll wager that something similar was seen in Nokia shortly before they were wiped off the commercial map.

Wisdom is understanding the limitations of the numbers alone: however they are crunched. Wisdom in business is understanding that it is not weakness to embrace wider ideas. Wisdom is strength, and this does not just come from data alone. Ultimately, wisdom comes from within, but the insights and context makers should be part of the mix.

If you are struggling with a business problem, or drowning in data without context, and you suspect that having a deeper understanding of how data works would help then please connect to me on LinkedIn with a brief message and we can start a conversation. Thank you.