Having spent most of my career as an analyst in the fields of Finance and Enterprise Technology, the word “data-driven” has become a security blanket of sorts. It’s the prefix we attached to our decisions to provide some measure of comfort that we weren’t the irrational human beings prone to bias and unrefined heuristics we really were. Most of my days were spent collecting data the same way my mom would collect free samples at grocery stores and malls (“I’m not sure what I’ll use it for but it’s free so take it!”). My fellow graph pushers and I would attempt to pull some interesting facts from the numbers, doing our best to remain objective, statistically significant, and tasteful in our choice of chart colors. Our powerpoints and excel reports would then get shipped out to a cabal of overlords who, mostly behind closed doors, would make “data-driven decisions”. I’m not convinced they did though.
I’m willing to bet that while data may indeed have triggered a discussion and may even have formed the basis of a list of possible actions to take, the decisions themselves are shaped by multiple, almost invisible factors. Such factors are inherent costs to group decision-making that must be paid for upfront as careful consideration or disregarded resulting in sub-optimal outcomes (Mark Curtis lays out a few such costs in his blog post). Consider the following the next time you’re about to make a group decision:
- People who weren’t able to attend a critical meeting due to scheduling conflicts whose feedback would have altered a decision
- A person – usually an older white male – whose words carry more weight simply because of who they are
- Insights being watered down to prevent a leader’s ego being bruised or a potential inter-departmental power grab from erupting
- The struggle to keep discussions on track due to the abundance of red herrings, tangential points, and – in the case of online chats – cat videos
- Guessing whether decision-makers actually studied your carefully written report in detail or if they merely glance at it before choosing the path they’ve already set their minds on weeks ago based on their “decades of industry experience” or molding your narrative to fit theirs
The collection and analysis of data does not a data-driven organization make and we often forget that there exists a large chasm between the analysis and the decision; a valley open to interpretation. Why does any of this matter? Because we have come to a point where the decisions of a few now have the power to impact countless others from individual influencers spreading false information about COVID-19 vaccines (or the disease itself) to a small number of meatpacking plants holding entire food supply chains together. Our complex global problems carry with it an increasing need to harness our collective intelligence while accounting for decision costs such as bias, geographic boundaries, and disparate data points.
Take food for example. We cannot presume that moving to a mostly plant-based diet in one region can work in another especially if the latter has a younger, poorer population living in a more arid environment and making policy changes in one place affects everyone up and down the supply chain and across oceans and mountains. We need to harness the intelligence of the cocoa farmer in Ghana, the fish processor in China, the agronomist in the Netherlands, and the diner in America. And we need to harness it fast and devoid of “don’t-read-the-comments” ire.
I can’t say there’s a quick solution to the high costs of our decisions though people are working on it. But an awareness of the costs is a good place to start.