Research has suggested that we overestimate our accuracy by about 20% when left to our own devices when dealing with a complex decision where we do not have deep experience or insight.1 Thankfully, access and capacity to increasing volumes of data is helping to overcome this decision bias.
In a recent survey by Cap Gemini is was determined that “participants estimate that, for processes where Big Data analytics has been applied, on average, they have seen a 26% improvement in performance over the past three years, and they expect it will improve by 41% over the next three.“
This is good news for the data industry…but I am a data skeptic. What does that mean? It means that I don’t trust data, or more specifically I do not trust data alone.
Often, data is positioned as the antidote to an overconfidence bias, but the truth is that data itself can be one of the drivers of overconfidence as Jim Harris points out in his post “Big Data and the Treadmill of Overconfidence”
In an interesting twist, a McKinsey report highlights the fact that for leaders who need to get things done, “Holding a somewhat exaggerated level of self-confidence isn’t a dangerous bias; it often helps to stimulate higher performance.” 2
In other words, overconfidence could be just the thing we need when making decisions!
In short, we can’t rely on data alone to drive a good decision
To make good decisions we need a lot more than good data and good analytics, and we certainly can’t rely exclusively on our own over confident intuition. We need good Decision Systems, and by Decision Systems, I mean more than the currently popular logic based Decision Models, If A, then B that are algorithm driven and real time.
I mean holistic, systemic decision models that include people, and process, biases and other slippery constraints, and which are the underpinnings of the majority of real decisions made today, in the past and probably for the foreseeable future. Decision Systems that take into account psychology and organizational behavior
In our next post, we will tackle the key components of a good repeatable, self learning decision system, and how data analytics provides a measured improvement for our clients.
What has been successful with Decision Systems you have used? Share your thoughts @firststepsinc
- Lichtenstein, Sarah; Fischhoff, Baruch; Phillips, Lawrence D. (1982). “Calibration of probabilities: The state of the art to 1980”. In Kahneman, Daniel; Slovic, Paul; Tversky, Amos. Judgment Under Uncertainty: Heuristics and Biases. Cambridge University Press. pp. 306–334. ISBN 978-0-521-28414-1. ↩︎
- http://www.mckinsey.com/business-functions/strategy-and-corporate-finance/our-insights/the-benefits-and-limits-of-decision-models# ↩︎