preparing-for-the-coming-data-tsunami-myth-busting-and-predictive-analytics

Preparing for the Coming Data Tsunami: Myth Busting and Predictive Analytics

“Trust the data.”

How do you react when you hear someone speak these words? Are you ready to believe what’s coming next, or are you on your guard?

In my thirty years of conducting data analytics for all types of people and organizations, I’ve learned that it’s better to be pragmatic than ideological. The ideology of data holds up numbers as truths. In reality, data is a practical – and extremely useful – tool. As we watch the crest of a new wave in our current tide of data enthusiasm, I recommend taking a pragmatic stance: You’re going to need it.

Predictive analytics comes with the promise of providing great insight and knowledge for decision-making. But as advocates of data, we run the risk of trusting and believing when we should still be digging in and questioning. Data’s new tools, buzzwords, and evangelists won’t change these four myths to look out for.

Myth #1 – Data is objective.

I heard these words pass through the lips of a Wall Street Journal reporter at the SXSW Interactive Festival in Austin, Texas, as he proclaimed that data is neutral. Unfortunately, data isn’t any more objective than a reporter. Like the facts of a story, data becomes filled with emotion when it’s selected and interpreted by someone else. How the data is collected, how it’s processed, and how it’s visualized are all human decisions. Data can provide a frame of objectivity – but don’t default to the view that it is.

Myth #2 – Data speaks for itself.

Data speaks the language of computers. Computers know how to process data and make it mean something. Most humans don’t have those skills. Where computers make sense with numbers, we make sense with stories. So that means someone has to interpret, and give data a human voice in order for it to be heard. That’s why we’re hearing about “storytelling with data.” People recognize the power of data, but it still needs to be translated into a language humans can understand.

Myth #3 – You can trust data more than you can trust people.

Learning who to trust is a hard-won lesson. It’s the same with data. With its practiced air of respectability and objectivity, we want to believe that data has our best interests in mind. It looks clean and acts precise. We assume it’s telling the truth about where it’s from and what it does for a living. But we have to learn the critical thinking skills – knowing what questions to ask and picking up on red flags – before we can put stock in any dataset.

Myth #4 – Data can strengthen your case for you.

It’s called cherry picking the data, and it’s all too common. It’s the practice of working backwards from your conclusions to find data that will strengthen your argument. This has given data a bad name. Even worse, it gives someone the excuse to trot out the tired joke about the three kinds of lies: Lies, damn lies, and statistics. Don’t be a cherry picker. The really smart consumers of data will see right through you, and you’ll lose their trust.

Over the next few years we are going to see these myths resurface again and again, with new words and new people. Data is getting smarter. It’s learning to learn. That’s why we have to be more pragmatic than ever before about data, and everything that comes along with it.

Predictive analytics use existing data sets to identify patterns and forecast trends into the future. Machine learning is getting a system to teach itself with data, instead of following pre-programmed set of rules. Artificial intelligence – for decades, you’ve heard this term – is here.

Companies are already using it. How does software know how people are going to respond to your speech before you speak? Quantified analytics firms can tell you – through predictive analytics – how an audience will receive your speech before you even say a word. Social media is providing a vast and rich trove of data, which businesses mine to target their ads with unprecedented precision.

I’m well aware that giving machines the ability to learn is the basis of many dystopian fantasies. I can’t deny the moral and ethical challenges presented, but the transformation is inevitable.

The new world of predictive analytics brings better ways of knowing, but it will also be more complicated, with less transparency than we have today. New myths will be created. Data will be smarter, but we can get even smarter – and take advantage of the benefits without choosing an ideology.

Peter Zandan, Ph.D., H+K Strategies, Global Vice Chairman – Research, Research and Data Insights, Global Chairman