Kathy Gibson reports from CeBit in Hannover, March 2016
There can be no doubt that we have become a data-driven society – but the way we do analytics today is arguably wrong, having already led to some major disasters and more on the horizon if we don’t learn to do things differently.
This is the warning from Christoph Holz, MD of Visalyze, who says we have to learn a new set of survival skills in order to thrive in the data-driven society.
“We are driven by data: but how can we become the driver again?” he asks. Setting the scene, Holz points out that global communication has become a basic need, and is changing the way people interact with one another “Digitalisation changes us: we become what we behold,” he says. “First we shape our tools, then the tools shape us. But what will this new shape be and what might be lost?
“What skills will we need in the future and what cultural techniques will we need in the information age?”
One of the main problems with the way we create, analyse and consume data is that our human creativity is rooted in visual thinking.
“This capacity of our brain is very old; hundreds of thousands of years older than our capacity to talk and write,” says Holz.
For instance, he says, most people remember faces better than names, and could be good at spatial orientation but bad at algebra.”
In the network age, the number of connections grows exponentially. The network effect states that with one phone, no-one can connect; with two phones, two people can connect with each other; but with six phones, there are 50 connections.
These networks are not evenly distributed, though, and calculating paths through them can become very complex very quickly. And the pace of change is not slowing, but increasing exponentially as well. “From the invention of the steam engine growth has been exponential and this is still true. For most of history, knowledge growth has doubled every century; after World War Two, it started doubling every 25 years; and today it take only 30 months for our knowledge to double. There is a new invention every three minutes.
“How can we cope with this?” Holz asks. “We cannot learn in this exponential curve?”
This leads to the subject of big data, and Holz invites people to ponder what big data has to do with us, and how we should be using big data to make the right decisions.
“Big data is not only about volume, variety and velocity,” he says. “It is mainly about veracity. You cannot trust your resources anymore; and that is why analytics won’t work any more.”
Holz believes that using visual metaphors would be more useful in discovering the value in big data, and helping companies to act upon. “Because all analytics is useless if you do nothing with it,” he says. “And I don’t think this is something that should be left to machines.”
For instance, he says the way we measure economic metrics is flawed and could go badly wrong. “The sub-prime crisis was caused by complex financials linked only by KPIs – and we believed the information we got. Mark Twain warned us about lies, damned lies and statistics – and this relates to KPIs as well.”
In conventional analytics, says Holz, statistics strip away much of the richness from the original data and arrive at KPIs, which could be incorrect because they don’t access all of the original information. “So what’s the next step in data management?” Holz asks.
He believes that visualisation is the key. “We all have mental models in our minds. These ae predictive and explanatory for understanding the laws of interaction. It’s how you are able to drive your car; how air traffic controllers can land aeroplanes and surgeons perform operations. “Understanding is about more than the numbers. It’s also about intuition.”
Traditional business intelligence only gives answers to the questions it is asked, Holz points out. “You won’t find the anomalies this way. Data visualisation, on the other hand, is about finding the questions and the answers can follow later.
“It’s important to remember that 90% of our sensory input comes from the eye; and visual processing takes up 50% of our brain capacity.”
The industry needs to find the next big thing, he adds: the mass market killer app for the data driven society that everyone will want and be willing to pay for.
“Gartner gives us a hint: at the top of the hype cycle is advanced analytics self-service. I think that people will start doing their analysis yourself – after all, you are the domain expert and you can do your own analysis. The next thing after that will be self-service for citizen data science. Data has to be managed and self-service will see the re-introduction of intuition and action to data analysis for citizen data scientists.”