by Margaret Lisi
Big data, never far from the news these days, has been taking a lot of hits on credibility, ranging from Facebook’s ad server presumptions about what we’d like to see to Microsoft’s chatbot spewing racist garbage. But inarguably, the biggest Big Data misstep this year was the presidential election.
I’ve wondered at others’ fascination with big data — corporations I’ve worked with seem to see it as a Holy Grail of sorts, with all the answers they need to cook up successful marketing plans and rabid brand fans. Big data is seen very much as a science in which businesses should invest millions of dollars and the futures of their companies, but here’s the caveat: there is an art to that science. As companies become more comfortable moving toward huge decisions based on the power of big data, a little voice in my head says, “yeah, but what about the people?”
Because, as author Mary Roach said, “People are messy, unpredictable things.” As marketers, we naturally rely on data to tell us big picture things, like buying and market trends, demographic breakdowns and more. But as the election showed us, even within a stated demographic, people can be very different and you paint them all with the same brush at your peril.
In my mind, people should be the first filter we use before making a big business decision. We certainly have the tools at our disposal—our ears, our feet, and our eyes. People can hear both what’s said AND unsaid, and get a first-hand sense of how the wind is blowing. Dr. Wolfgang Heuring calls this “smart data.”
New York Times reporters Steve Lohr and Natasha Singer wrote in their post-election story, How Data Failed Us in Calling an Election, “It was a rough night for number crunchers. And for the faith that people in every field — business, politics, sports and academia — have increasingly placed in the power of data.” And it should have been, because even as our data has gotten better, people are still central to the equation. People read and interpret data, and not everyone is an A-player at data analytics (not knocking you folks at the NYT).
In the same New York Times story, Dr. Pradeep Mutalik underscored that big data and our increasing reliance on it is “the overselling of precision.” We’ve come to see big data as gospel, when the most unpredictable element in any equation is x, the unknown “people” factor. Despite the most sophisticated algorithm, the only reliable thing about people is they don’t do what you predict they will 100 percent of the time.
CNN spoke of Republican voters sticking with Trump as though we were lemmings. I registered Republican when I was 18 and my dad had a great deal of influence over my political views. It’s much later now, and I am not loyal to that little R on my voter registration card, preferring to vote depending on the election and whose platform makes the most sense to me. That makes me, in a statistical sense, X. I am a wild card, and there were a lot of us Nov. 8.
People are complicated. They may zig and zig and zig, then zag when you least expect it. Even the “smartest” software can’t predict emotion, or feeling invisible, or any other situation that results in a piece of data someone, somewhere, is “crunching,” as though we’re interchangeable building blocks without shade or nuance. In fact, that number cruncher is often so far removed from the origin of that data that they can, and do, draw the wrong conclusions.
In fact, South Dakota entrepreneur Neal Tapio said he saw and heard the electoral tide turning for Trump long before election day in his working class town of Watertown, South Dakota, simply by talking with folks who were scraping by, living lives unfulfilled by prior political promises.
In business’ effort to get ahead of its competition and make more accurate decisions, I see us getting away from the incredibly important “people” part of decision-making: having real people interacting with your customers, employees and partners, then layering that qualitative, human perspective on top of the numbers.
I believe this year’s election may prove to be a powerful turning point for Big Data and those who believe its gospel. I believe big data may tell you the stories you look for, often due to analysts’ or leaderships’ intrinsic bias to find numbers that support their particular view. This election, Trump’s data team lead Matt Oczkowski may have been one of the few who saw Trump’s win even as almost every other analyst assured voters that Clinton had the presidency in the bag.
Data analysts are people too, and so can be intrinsically biased. Without people you trust on the ground and in the field, telling you what’s really happening in your business so you can add that layer to your data dump, you may miss the real story. –That’s not a risk I’d want to take with my business—how about you?
headline image created by izquote.com
Partner with us.
Let's tell your story and drive the change you want, today.Say Hello