What do dirty, worn sneakers have to do with turning around a business?
If you’re LEGOs, it means everything.
LEGO executives – who had been told countless times by experts that they were in trouble because future generations would lose interest in the building blocks – decided to pay an 11-year-old boy a visit in 2004.
The boy was not only an avid LEGO user, but was passionate about skateboarding. He told LEGO executives that his most prized possession was a pair of ratty Adidas shoes, showing the executives how the sneakers were worn in just the right areas. That was key, the boy said, because it proved that he was one of the best skateboarders in the whole city.
The light bulb went off for the LEGO executives. They saw immediately that this boy wasn’t only interested in things that were quick and provided instant gratification like computer games. Instead, it was the social currency among his peers that mattered most, and he was willing to put in the time and effort to achieve a high level at something.
As author Martin Lindstrom explains in his book, “Small Data: The Tiny Clues That Uncover Huge Trends,” LEGO’s decision making until that point was based entirely on “reams of big data.”
But once they had the encounter with the boy and his beat-up sneakers, they began refocusing on their core product, moving away from the bigger LEGO bricks and going back to the original, smaller size. In fact, the company added even smaller bricks, more detail, and instructions that were more exacting. Overall, LEGO projects became more labor intensive.
LEGO executives had learned through their interaction with the boy that for users it was all about “the summons, the provocation, the mastery, the craftsmanship, and not least, the hard-won experience,” a conclusion totally missed by their complex predictive analysis, Lindstrom says.
LEGO, he says, is not the only company to be blinded by big data.
“Big data has convinced board rooms across the world that they’re on top of what’s going on. Nothing could be further from the truth. It has become fashionable to refer to big data without questioning its true value,” Lindstrom says.
What is being missed is the “small data,” which can yield critical clues, he says.
“Big data searches for correlations of data; small data identifies the actual causation behind the data. You can’t begin to draw correlation before first identifying the causation, because the causation almost always points to a larger context—frequently shedding a very different light on the situation and indicating what one should really search for,” he says.
For example, in 2012, Google concluded that – based on search terms—it could predict a flu outbreak days before it would happen. Doctors and pharmacists across the country would be able to order pharmaceutical products well in advance, Lindstrom explains.
“This was deemed revolutionary,” Lindstrom says. “Yet just recently, the Center for Disease Control revealed to everyone’s surprise that the data from Google was twice what it should have been. They found that when we begin typing searches like ‘flu,’ people around us will follow us, curious about what’s going on—resulting in a misinterpretation of the data.”
In other words, Google had focused on identifying correlation, forgetting about causation, he says.
“Every day, large corporations base their entire futures on big data, assuming these are based on solid data. Yet as they come to rely on these sets of data, organizations are slowly moving away from the consumer and true market conditions. A counterbalance is needed, and the answer is small data,” he says.
Lindstrom advises that leaders use small data to identify the hypothesis, while using big data to help confirm the correlation since like “partners in a dance the two sources are perfect at verifying each other.”
In addition, small data may ease the way for non-tech workers who may be intimidated by big data and avoid using it.
“As small data is based on human interactions it is in my experience substantially easier not only to obtain these data but also to interpret these,” he says. “Sure it – as everything – takes time to adapt to a reality where the staff will have to interact with its customers on a regular basis, yet once this routine has been embedded into the organization it is my experience that the company becomes remarkably more nimble and consumer focused.”
The importance of spending time with consumers should be underscored by leaders, he says, making it “just as acceptable as reading consumer reports” and rewarding those “who pride themselves in observing and spending time with consumers.” Further, sharing these results with others in various functions within the company will help change the mindset to one that is more customer-centric and help break down silos within the organization, he says.
Lindstrom says that recently one of the largest food companies adopted a policy to ensure that almost all its staff worldwide must visit consumer homes twice a year. “That’s a testimony to the fact that there’s nothing more powerful than witnessing and experiencing a consumer’s life first hand,” he says.
He says it’s time that companies took a deep breath and reassessed when and how they use big data.
“Small data are human. I mean none of us would ever describe our loved once using a spreadsheet – yet the persuasive power of numbers and statistics often lure us and senior management into believing that true emotional definitions can be described in details via conventional big data,” he says. “It is however becoming more and more evident that true emotional insight needs to be found by first hand observations—leading to powerful hypothesis which again will create the foundation for further big data mining.”