The value of data in decision making is getting a lot of attention these days. Test everything. Collect data. Examine the data and see what it tells you, because, after all, data can’t lie.

Not so fast.

For years experts have been telling us to stop listening to our gut and start paying attention to data. But now the experts have changed their collective minds and are telling us that our guts still have something to say.

Part of the problem with depending too much on data is that it’s isolated. “There is so much invisible, fluid context wrapped around a data point that we are usually unable to fully comprehend exactly what that data represents or means,” writes Dan Zambonini in Medium.

“We often think we know, but we rarely do. But we really WANT it to mean something, because using data in our work is scientific. It’s not our decision that was wrong—we used the data that was available. Data is the ultimate scapegoat.”

Zambonini goes on to cite the song “Bohemian Rhapsody” by Queen, which, according to data, should never have been successful: It’s too long, too complex, too confusing, and it didn’t fit into any musical genre, according to radio stations at the time. Yet it reached the number-one spot in four different years. And admit it, you’ve sung along to it one time or another.

Zambonini is particularly interested in this from the context of startups, which he sees as being overly reliant on data rather than on vision. “I hypothesize that an early startup guided primarily by gut decisions from a strong strategic vision will be more cohesive and deliver a stronger offering than a startup created from a random walk of data-driven decisions.” (Zambonini admits he doesn’t have any data to back this up.)

“Imagine what would happen if Apple had relied solely on customer data and input when they were creating the first iPod, or the iPad,” writes Dawn Hepper, former vice president of strategy for Ciceron. “No data was going to tell them they needed to create these products.  They had a vision, an instinct, and they went for it. And it paid off.”

This is also true in the case of design, such as the example of Douglas Bowman, Google’s visual design lead. “When a company is filled with engineers, it turns to engineering to solve problems,” he wrote in a blog post explaining his decision to leave the company in 2009. “Reduce each decision to a simple logic problem. Remove all subjectivity and just look at the data. Data in your favor? Ok, launch it. Data shows negative effects? Back to the drawing board. And that data eventually becomes a crutch for every decision, paralyzing the company and preventing it from making any daring design decisions.”

Examples he cites include testing 41 different shades of blue and having to justify decisions on whether a border should be 3, 4, or 5 pixels wide.

Design, in particular, also contributes to more than just the color of the button or the width of the border, but to the company’s quality and brand, writes Braden Kowitz in Wired. “A good reputation takes years to build, but just one bad experience can destroy it,” he notes. “So when balancing between tactical easily measurable goals like more clicks, and long term goals like trustworthiness, it’s essential to listen carefully to one’s instincts.”

Some research has even indicated that gathering more data makes people’s decisions less accurate.

That’s not to say we should throw big data out with the bathwater. In fact, a number of trust-your-gut proponents believe that the issue of intuition vs. data is a false dichotomy. What’s really needed, they say, is to find ways for intuition and data to work together.

“It’s foolish to assume that number-crunching alone can provide answers a business needs to get ahead,” writes Steven Hillion, co-founder of Alpine Data Labs. “In data science, intuition and analytics work together in tandem, each informing the other.”

For example, an engineer working for an aerospace engine manufacturer cannot expect diagnostic software alone to determine the causes of problems, write Donald A. Marchand and Joe Peppard in Harvard Business Review. “Rather, the engineer must have considerable expertise and knowledge to identify relationships in and ask questions about the data, often through the testing of hypotheses. And in interpreting the results of any analysis, he or she must draw on experience to weed out misleading or false explanations.”

While Zambonini speaks specifically about startups, what he says could be applied to any company. “I don’t believe that an early startup should be driven by data,” he writes. “It should be assisted by data, using it as one of a number of inputs when making tactical decisions.”

This is particularly true when the data contradicts intuition—and, in fact, many executives say that when such contradictions happen, it leads them to go back to the data and look at it again rather than tossing aside their intuition.

In one example cited by Toby Wolpe of ZDNet, executives found that a particular test insert had no effect on customer behavior. Instead of dropping the insert, they went back and discovered that the test had actually never been performed. Oops.

Related Posts