3 ways data analytics can fail

Whether you are using a spreadsheet or advanced data analytics with machine learning or AI - unless you are vigilant, your foraging for insights from data can so easily go wrong.

A big part of the problem is that ultimately, data is always viewed through a human lens. Unfortunately, our brains did not evolve to process complex numerical data and our instincts, biases and desires can significantly alter our understanding of the data being presented to us. We have no way of knowing if data is right or wrong simply by looking at it. So we have to be very careful when making decisions based on it.

In this short video - 3 causes of insight failure are highlighted. Each compounding the risks that our decision making may be flawed.

• GIGO - Garbage In Garbage Out
• People see what they want to see
• Lies, damned lies and statistics

I also include some hints at ways to avoid these problems.

Data Analytics - past, present and future

The desire to look into the future is as old as humanity itself. From wanting to know where the next meal is coming from to placing bets on the Grand National - a view of the future has an obvious attraction.

Nowadays, businesses are turning from witchcraft and wizardry and looking to something a little more scientific. They want to figure out what is to come and give themselves an edge. 

These days, it's all about data science.

New technologies like artificial intelligence, machine learning and cognitive systems, coupled with the promise of Big Data are creating the illusion that all questions can be answered - even those which pertain to the future.

The challenge of course is that the only data we have - is from the past. And whilst you may subscribe to the view that past performance predicts future performance - you will in fact be disappointed. It's just a question of when.

Of course, data science can, if wielded correctly, sustain a competitive advantage. It can swing the odds in your favour as you navigate an uncertain future. If your predictions are 10% more accurate than your competitors - well that is an edge. 

It's like the adage of outrunning a bear. In fact you don't have to, you just need to run faster than the person next to you.

To find out more about using data from the past to foresee the future and the role of data scientists - watch this video on Data Analytics. This is Part 1 - from descriptive to prescriptive.