As we settle down with our families to enjoy the winter holidays, many of us will be celebrating various traditions.
For many, the culmination of the season is waking on December 25th to the excited cheers of their children as they see their Christmas horde waiting under the tree. These children believe that Santa Claus, despite carrying what must be a massive load, manages to fly on a sleigh powered by eight reindeer to every home on the planet in one night. The jolly old elf climbs down the chimney and places each present carefully underneath the Christmas tree. The children wake up on Christmas morning and, ta-da, there are presents.
As it turns out, there is a word for this belief: Apophenia. According to Wikipedia, Apophenia is defined as the experience of seeing patterns or connections in random or meaningless data. Apophenia was first used to define the "unmotivated seeing of connections" accompanied by a "specific experience of an abnormal meaningfulness.” While more critical thinking minds, and older children, may question the entire Santa process, the far easier approach is to accept the data given the outcome.
In business, this human tendency to seek patterns in random information can have misleading, if not disastrous outcomes. It often is a result of bad data analysis – someone wants to see something so they see it. It also occurs because of bias in the data itself. We’ve written about that recently in a post titled Big Data Hype and the Parable of Google Flu. In this now famous example, Google misreported that flu levels reached 11 percent in the US winter 12-13 flu season, almost double the CDC’s estimate of about 6 percent, based on its search data.
To avoid human and data bias in business, it is critical to understand scientific process and to improve the flow of data, analytics and the operationalization of data insights throughout the business. Elimination of bias should occur not only upstream – what are your data sources, who are your users – but also downstream – understanding how and where data-driven decisions are being used and acted upon.
As more organizations codify their decisioning process and institutionalize their data from insights into action, they also need to eliminate human and data bias. Even in the creation of the data analytics themselves, the addition of a champion/challenger and machine learning components will often eliminate less than optimal performance due to human interaction. This isn’t to say that humans should ever be completely eliminated from decisioning process, but rather that the elimination of human and data bias can dramatically (and positively) impact customer engagement and business outcome.
While we must diligently avoid Apophenia in our business lives, it is perfectly appropriate for it to bring magic to our winter holidays. No matter what traditions you and your families celebrate, I wanted to wish you a Happy Holidays and a joyful New Year.