Every time I think of Big Data I reminisce back to my childhood and the Brothers Grimm tale, Hansel and Gretel. Individuals today are very much like this story, leaving trails of breadcrumbs as we live our lives. Marketers follow those trails to decipher who we are and the behaviours we carry on.
Marketers aren't the only ones examining the trail of data for clues. It's human nature to constantly look for patterns in data. When two sets of data are strongly linked together we say they have a high correlation. In its simplest form, it answers the question: does this quantity change when this other quantity changes?
Certain correlations just make sense, such as the strong correlation between warmer weather and ice cream sales.
Consider the following correlations. They aren't too surprising:
• In the medical field, studies often find a positive correlation between severity of illness and nutritional status of the patient.
• There is a well-known and proven correlation between smoking and lung cancer.
• Research has shown that the sales of gin correlate with sales of tonic water.
• In 2009, experts at Forrester Research uncovered a direct correlation between client or customer satisfaction and increased revenue. The impact of client satisfaction may increase revenue in some industries by as much as $1 billion annually.
Here are some other fascinating correlations:
• In 1992, a team of data scientists looked at point-of-sale data from a regional drug store chain and discovered that between 5pm to 7pm customers tended to co-purchase beer and diapers.
• A Harvard Business School study in 2011 determined that a one-star bump on Yelp leads to a big revenue boost.
• Recent studies by university researchers have proven a link between customers’ use of social media and higher revenues. Participating customers on a brand's social media site contribute to 5.6% more revenue, as well as 5% more visits to the business, vs. non-participating customers.
• Shopzilla observed in 2009 that faster page load times equated to a 12% increase in revenue.
• Published findings suggest that people increasingly rely on motorcycles to reduce their fuel costs in response to rising gasoline prices.
• Two independent research samples found that women are more likely to wear reddish clothing when at peak fertility – presumably as a way of increasing their attractiveness.
Beer and diapers. Who knew? That is, until the bread crumb trail was followed and the data revealed the pattern. Thinking about it now, the correlation between beer and diapers is pretty obvious – if you have kids, which not everyone does. But this is the type of correlation that represents “gold” for the savvy marketer, and information that can harness increased revenue, which otherwise would be missed.