How to out-argue a data guru at the office
Share on Linkedin
Big data can be useful, but it only tells part of the story. (Thinkstock)

How do you take a stand for intuition, creativity or just plain old critical thinking?

Quick, count the number of things wrong with this scenario:

Your company's new big-data guru takes the floor during the monthly brainstorming meeting to discuss a talent shortage. His latest graph shows a correlation between the number of outsiders taken to lunch by the staff, website visitors and new hires. Seduced by the data, the chief executive officer assigns everyone in the room, including the shy software-coding guy who hardly ever leaves his desk, to have three lunches a week with friends to find new prospects.

The marketing director slips her witty, innovative new recruitment ad back into her folder, unseen. Everyone leaves the room, wincing. Big data wins again against big ideas.

In the past five years, the ability to tap data has allowed us to access information that was never accessible before. But data has its limits. Some executives treat the new tools like shiny new toys, and apply them to problems data alone cannot solve, say experts.

Even data's biggest boosters are a little worried.

"The recitation of a statistic (sometimes) shuts down critical thinking," Nate Silver, the statistician and journalist famous for correctly predicting how all 50 states would vote in the last US presidential election, said in a recent speech to the Online News Association in which he warned of the hazards of ignoring data and using it blindly.

The world produces 2.5 quintillion bytes every day — enough to fill CD-ROMs stretched from Atlanta to San Francisco, according to Silver.

If you’re that marketing director or project manager, hiding your idea in a folder in the face of insurmountable numerical evidence, you may be wondering what you can do to head off the data train barrelling toward your ideas. How do you take a stand for intuition, creativity or just plain old critical thinking — even when the data disagrees?

Don't let yourself be sidelined.  At the end of the day, data is still just a tool, best used to describe the past, and, cautiously, to forecast certain carefully defined future events, said William Duggan, a management professor at New York’s Columbia Business School and the author of four books about strategic intuition. "The mistake is thinking data will give you the idea of what to do," he said. "That is not where ideas come from."

Good ideas are the synthesis of things that have worked in the past, Duggan said. The best synthesizers often are people trained to trust their intuition and creative instincts — the strategists, marketers, designers and wordsmiths.

A human-centred approach to solving problems — called design thinking — marks some of the greatest business successes, according to the new book Creative Confidence, by brothers Tom and David Kelley, general manager and founder, respectively, of California-based design firm IDEO. Apple, the maker of the iPhone and iPad, is a poster child for illustrating how successful projects can be when they tap creative spirits, the authors say.

Embrace the creative process and stand your ground. Creative processes like brainstorming and synthesis (combing old ideas into new concepts) have been shown over and over again to lead to success.

Use examples. The power of the human mind to detect and create patterns means that a well-placed example can sometimes shift the conversation, according to Duggan. To make a strong argument without relying solely on data, or to influence a data-driven decision, you’ll need to reference case studies that are relevant to what you are proposing.

Some examples: "This is similar to the way Amazon entered the art market", or "We don't want to be the Nook vs the iPad", or "Here's how IBM won in the Spanish market."

Be prepared with your own data to prove that non-data-driven wins in the past actually were successes. How many clients were added? What was the increase in market share? That's using data the right way, to describe a situation rather than prescribe a solution.

Re-focus the conversation on the problem. London-based professional services firm Ernst & Young uses lots of data, but it tackles issues using a three-part approach to get results. The approach includes account data management, analysis and decision, said Christer Johnson, principal. The company starts at the end of that sequence by defining the problem and decision that needs to be made.

"Big data is a lot like sex in high school," said Johnson, quoting a joke that has been making the rounds on big data blogs lately. "Everybody is talking about it but nobody is really doing it and even fewer people are doing it well."

Johnson, who leads the advanced analytics services in EY's Enterprise Intelligence Competency from a Virginia office, tells clients that the data can help them make sense of a decision, but "it's not your job to go make sense of the data."

Remind the room of the human impact. Decision-makers driven too much by data may forget there are people on the other end of the equation. The most well-known cautionary tale: US retail giant Target's use of customer search data to send baby-item advertisements by mail to women likely to be pregnant. Predictably, some families weren't happy that a giant company was privy to the news of their impending arrivals.

One question to ask: does this use of data benefit the clients and the company, or just the company? "If you ask yourself if (you’re) making (the clients) lives better, that's one way to objectively navigate that area," Johnson said.

Question the assumptions.  A person too deep in the data weeds tends to ignore what can't be measured. In statistical terms, he or she may wrongly assign a value of "0" to incalculable things. For instance, the humour in the marketing director's advert might have an incalculable positive impact on the company's brand.

Other human tendencies may affect the objectivity or comprehensiveness of data, said Silver in his speech, including the fact that people ignore data that isn't useful to them.

Then there's the biggie: correlation — as in the correlation between lunches, visits and recruitment — is not the same as causation.  Silver explains it this way: two things can have a relationship statistically, but not really be related structurally. That is, there’s no real connection.

The questions to ask, then, when the boss insists everyone take three lunches with potential recruits each week: "Do we fully understand the relationship between lunches out, visitors to the website and new recruits? Are they really connected?"

Ask the simple question: How will we implement? This basic query can be enough to restart some of the critical thinking in the room. Sometimes, statistics are so appealing that people in power just want to use them, but haven't thought through the sacrifices of time and money that will have to be made in order to implement a data-driven idea.

The time of that shy genius in the corner is likely better spent debugging than on awkward lunches. But until the reality of the number is separated from the reality of the task at hand, that’s not always easy for managers with data stars in their eyes to see.

We want to hear your examples of big data stifling workplace creativity. Join the Facebook conversation to share your experience — or your solutions.

Around the BBC