Here’s some insight from Kellogg professors on how individuals, teams, and companies can get comfortable with and get the most out of data analytics.
1. Leaders must also understand Analytics
First, let’s dispel the myth that you just need to hire a crackerjack analytics team.
Florian Zettelmeyermarketing professor, emphasizes the need for business leaders to be comfortable with analytics as well.
This does not require a degree in computer science. It takes what Zettelmeyer calls a “working knowledge” of data science. This means you can separate the good data from the bad and know exactly where analytics can add value.
Zettelmeyer says most managers share a common behavioral bias: when results are presented as achieved through complex data analyses, managers tend to defer to the experts. However, it is the manager’s job to choose which problems should be solved and how the company should integrate analytics into its operations. Executives are the decision makers, so they should play a central role in determining what will be measured and what the numbers mean for the company’s overall strategy.
Managers also have valuable knowledge about the business, which they should use to explain strange results. Zettelmeyer recommends asking the question, “Knowing what you know about your business, is there a plausible explanation for this result?”
2. Tips for creating an Analytics team
Of course, a crackerjack analysis team is also a plus. So how do you hire one, especially given that top-quality data scientists are often in short supply?
Data analytics experts offered advice during a session at the Kellogg on Growth Forum a few years ago, moderated by marketing professor Eric Anderson. Here are some of their tips.
Dan Wagner: CEO and founder of Civis Analytics, former head of analytics for the 2012 Obama campaign
“What I’ve learned from the thousands of people I’ve interviewed is that I don’t have the mental capacity to talk to a person and assess whether or not they’re going to be good at the job. What you can do is simulate the job for them through an examination process. See how they do and rank them against everyone else. What you’ll find is that introverts tend to do much better, and that’s the classic person you’re trying to hire.”
Leslie Hampel: Then director of global strategies for Starbucks and now vice president of stores
“What I’m looking for is someone who can bridge the gap. Can you do the math? This is important, but can you get the story out of the math? Especially for the next 10 years or so as we move through this current generation of CEOs who don’t understand algorithms for the most part.”
3. Create a culture of intellectual curiosity
The best companies create a culture that encourages using data to solve problems, which goes far beyond the analytics team.
“It’s about the encouragement, the anticipation, and the ability for people to say, ‘Hmm, I wonder how we could use data to predict it or improve it or optimize it?’ says. Tom O’Toolesenior associate and clinical professor of marketing at the Kellogg School, who was previously CMO at United Airlines and CMO and CIO at Hyatt Hotels Corporation.
Questions should be welcomed from all corners of an organization. For example, O’Toole describes a meeting with the senior leadership team of a financial services firm. The general counsel wondered aloud how data and predictive analytics could be used to identify potential instances of a particular type of regulatory compliance problem.
“People didn’t expect the lead attorney to be the one asking how to use predictive analytics to deal with an issue,” O’Toole says, “but he did. And anticipating and avoiding the compliance issues he had in mind would have real business value in avoiding the costs associated with lawsuits, customer settlements and complaints.”
Organizations should have an explicit expectation that employees use data in new ways to answer new questions. One way to do this is to define intellectual curiosity as a key criterion for progress.
“This isn’t just ‘How many interesting questions have you asked and answered in the last six months?’ says O’Toole. “This is more of a mindset assessment: How do you look at and advance the business in new ways using data?”
4. The Paradox of Analytics
Now your analytics team is humming along, solving all kinds of business problems.
But don’t get too comfortable, say Zettelmeyer and Anderson, who are academic directors of Kellogg’s Executive Education program at Leading with Big Data and Analytics and they are writing a book on data science for leaders.
They warn leaders to be aware of the ‘Paradox of Analytics’.
The better a business becomes at gathering information from analytics—and acting on that information—the more streamlined its operations become. This in turn makes the data resulting from these operations more homogeneous. But over time, homogeneity becomes a problem: variable data—and, yes, mistakes—allow the algorithms to keep learning and optimizing. As the volatility in the new data shrinks, the algorithms no longer have much to work with.
The paradox leads to a rather surprising recommendation: “occasionally you need to intentionally mess things up,” says Zettelmeyer.
In some ways, the value of big data lies in its messiness—in the often unexpected variation in how events unfold, and in the myriad ways in which those events help make connections between variables that can help people get better decisions.
“Theoretically, the best manager for analytics is the one who comes into the office every morning and flips a coin to make all the decisions,” Anderson says. “Because if you make all your decisions by flipping a coin, you’re going to generate the best possible data for your analytics engine.”
Rather than flipping a coin, leading companies have adopted a strategy of injecting volatility into their data strategically.
A company developing a national advertising campaign, for example, may decide to modify the campaign in significant ways only in select markets or to scale the rollout by region. While there may be short-term costs in terms of efficiency and optimization, the resulting data has the potential to teach the company going forward.
5. Display your data
Armed with amazing data insights, how do you continue to get people to take action on them? Present your data visually, says the Northwestern psychology professor Steven Franconeriwho is also a professor of leadership at Kellogg.
“People might think visualizations are pretty and cool,” he says. “That is not true. They are necessary, absolutely necessary.”
Because our brains use the same systems to process speech and writing, putting text on a presentation slide while we’re talking “makes sure you don’t get your point across because no one can read and listen at the same time,” he says . .
But, he adds, “you can look at pictures and listen at full volume at the same time.”
However, very few people learn how to create good data visualizations. And it comes at a cost.
“If you get it wrong, it’s a disaster,” says Franconeri. “It could be whole spaghetti.”
Here are some tips to avoid spaghetti.
If you have a relatively simple data set, create three or four visualizations.
“Make them as different as possible,” says Franconeri. “Show them to some colleagues. Ask them what story they see in each. You’d be surprised at the differences in what people extract from the same data.”
With a more complex data set, keep in mind that your viewers can quickly become overwhelmed even if you fully understand the visualization. Because the visual processing centers in our brains are so powerful, it’s easy for us to become completely comfortable with a complex visualization because we’re immersed in the data. Young viewers, however, struggle without that background.
“Their brain is on, but they don’t know what to do with it all,” says Franconeri.
