Skip to main content

Why Your Analytics ROI Is Low (It’s Not What You May Think!)

It’s a familiar story. You’ve tried almost every combination of tips and tricks to improve your data analytics, but your Business Intelligence and analytics ROI is low. First the focus was data mining, then Big Data, and now it’s data lakes. But what’s the best way to boost data analytics ROI without wasting more time and money? The answer could be less about a technical solution and more about the nature of human beings.

This problem can be summed up very simply with a quote from a Nobel Laureate, Daniel Kahneman: “No one ever made a decision because of a number. They need a story.”

If you think about it, it makes sense. Charts and tables should be supplemented with an explanation, and the results of analysis need to be explained in order for everyone in your business to be able to make decisions from the data. Kahneman, along with being a Nobel Laureate, is a psychologist and one of the leading thinkers in the area of decision analysis. In his book, “Thinking, Fast and Slow,” Kahneman states that humans make decisions based on two “Systems”:

  1. an automatic, instinctual thought process that happens quickly (basically a fight or flight mechanism); and
  2. a slower, more deliberate thought process (a process based on reasoning)

His basic premise is that our instinctual side often overpowers our slower, more deliberative thought process. This happens when we are overwhelmed with data or when we don’t completely understand something. This simple, yet novel idea seems quite familiar to many, because who hasn’t heard a CEO say “I’m going with my gut” in response to numbers?

So if your data analytics ROI is low, the answer might not be a data lake tool, a Big Data solution, or a new analytics tool. The answer could lie with the need to explain insights in plain language. But how can you automate this process? Do you need a Natural Language Generation (NLG) solution to automate these descriptions? Our ebook looks at the history of analytics and offers a simple three-step test to see if NLG can help.