Analytics and Conversion

Decision-Making with Data (How Your Intuition Could Be Wrong)

How do you make decisions? Do you base your decisions on data or do you call upon your intuition? According to a survey by the Economist Intelligence Unit, “73% of respondents say they trust their own intuition when it comes to decision-making”. Trusting your intuition may be advisable in certain situations, but in others, your intuition could be wrong. As Boston Consulting Group founder Bruce Henderson eloquently put it: intuition is “the subconscious integration of all the experiences, conditioning, and knowledge of a lifetime, including the cultural and emotional biases of that lifetime.” Intuition is personal, it is your perception and it can certainly be a misconception proven wrong by data. 

First, let’s look at some specific examples where intuition have been proven wrong.

Unlike most people’s intuition pink is not the preferred color by women- it is blue. Despite what you imagine when you think about gamers, almost half of them are women and the average age is 35 years old. People think of Africa as a continent stricken with poverty. However, it’s not the case country by country - Botswana is a middle-income country and it is ranked in the 28th place out of 167 countries on the Democracy Index. In a survey done by Ipsos Mori, people’s perception of their country’s unemployment rate is 3-4 times the actual. 

If intuition can be so untrustworthy, what can we do instead? The short answer: data-driven decision making.

Data can help you make surprising un-intuitive business decisions. During hurricane season, the items most in demand are, unsurprisingly, flashlight, batteries and bottled water. Stores rush to stock up on these items when a hurricane is about to make landfall. The teams at Wal-Mart took it a step further. They mined datasets of shoppers’ history before another hurricane and discovered some surprising insights. They found that strawberry Pop Tarts increase in sales seven times their normal sales rate before a hurricane and that beer is the number one selling item pre-hurricane. Armed with this knowledge of customer purchasing behaviour, Wal-Mart was able to supply products that satisfied better the customers’ demands.

And for one movie studio, social media data helped uncover surprising insights about their own audience.

The movie Pitch Perfect was a cult hit that became the second highest grossing music comedy film of all time, and preceded the even more successful Pitch Perfect 2. Inspired by the success of the Glee TV series, the film portrays a college a cappella group who prepares to go to a national competition. A film where all the main characters are female, Universal Studio had been expecting an audience of mostly female college students who were fans of the show Glee. After consulting with the social media analytics company Fizziology, which mines social media data for insights, they discovered something unexpected. The Twitter data showed that most of the positive conversations around the film came from men. Armed with this knowledge about their own audience Universal redefined the film’s marketing strategy.

Perhaps nothing is more definitive of data-driven decision making than A/B testing, a method widely employed among top Internet businesses to improve their products. Using A/B testing, companies can test in real-time a redesign of a web page to a subset of their users, tracking the performance of this test, and comparing it to the control group. Here, any intuition about the effectiveness of each design change is backed up by the actual test results. 

No company is more data-centric than Google. Remember the story when former Google VP Marissa Mayer asked her team to test 41 shades of blue to see which one performs better on links to ads in Google Search and Gmail? You might think this move went a bit overboard, but according to Google’s UK managing director, choosing the right blue color helped boost Google’s revenue by an extra $200 million.

Another example when data proved intuition wrong is in regards to President Obama’s 2008 campaign. His team believed unanimously that a video of Obama speaking at a rally, placed on the home page of his website, would outperform any other still photo, prompting more users to sign-up on the site. However, after conducting an A/B test, they were shocked to find out that the video actually performed worse by 30.3 percent. Had they used the video, they might even have wrongly concluded that Obama was losing support of the public. With improvements from the A/B tests, the team measured that sign-up rate increased by 140% and an additional $75 million was raised.

Data is the only objective measure that provides a fact and evidence-based approach to decision making. Decisions doesn’t have to made by HiPPO:”highest-paid person’s opinion”. Instead of spending time debating the best font to use on an ad, you can instead A/B test it. Before making a decision, fact-check your assumptions by examining the data. Keep an open mind to learning new insights from data, by performing exploratory data analysis.

In conclusion, I leave you with words from the renowned American statistician W. Edwards Deming:“In God we trust, all others bring data”.

This article was originally written by Lucy Wang.

comments powered by Disqus