Commander Data

Star Trek: Next Generation was always right in my wheelhouse. I loved the optimistic and technologically advanced view of the future along with the adventure inherent in exploring the unknown. A favorite character was Data (the first to be discovered but second to be created, for aficionados): sentient android or artificially-created being through which the show often wrestled with the concepts of what it means to be human, logic versus social comprehension, the advantages and disadvantages of emotion, and their impact on decision-making.

I remember struggling early in my career with insurance business decisions in an uncertain environment.  No matter which path you took, there was always countervailing data to raise doubts about a decision.

An early boss kept advising me to rely more on “gut feel”. It wasn’t logical, but I learned that there was truth to it. We live in an era of massive data accumulation, processing and mining. Merrit Strunk, our Chief Marketing Officer, can wow you with how much data can be acquired about insurance agents – about your clients – for use in marketing and sales. However, data alone is often insufficient for decision-making, even if you have a lot of it, because there is often more unknown or unknowable than there is data.

David Brooks writes an interesting column highlighting some of the shortcomings of relying too heavily on “Big Data.” The human analysis that comes with reading people’s emotions, that depends on the feelings that surround relationships, that values non-financial or material items, like honesty, integrity and loyalty, is opaque to Big Data. We stay with certain relationships beyond their point of coolly weighing their pros and cons. That’s not a bad thing necessarily, but data alone won’t decide that outcome.

After an unknowable point, more data doesn’t bring us any closer to a certain conclusion; in fact it can easily confuse us, creating an unlimited number of apparent correlations where there is no real cause-effect. Just look at all the wrong conclusions health scientists have drawn from voluminous data. One example: Cholesterol is bad, vastly correlated with heart disease. Wait, the latest is that it’s a symptom, not the cause. Omega-3 / Omega 6 oil imbalance is a big cause of vascular disease by causing inflammation.1 Correlation doesn’t prove cause; it’s a wink in the right direction, but is insufficient alone. Even worse, data is neither raw nor objective. The old saw, “Figures never lie, but liars figure,” is half right. Data sampling and assembly is usually colored by the motivation of those doing it. Check out political polls.

You can trust that we at Creative, in our day-to-day dealings with you, make decisions with significant weighting to the social and values aspects of doing our business as those of you with us for a long time can attest. Brooks says, “Trust is reciprocity coated by emotion. People and companies that behave well in tough times earn affection and self-respect that is extremely valuable, even if it is difficult to capture with data.” I’ve learned that in 16 years at Creative: so true.

Join the conversation