7 Deceptions of Decision Making - An Analyst's Guide
Some cognitive challenges business leaders face while taking key decisions to drive their business objectives include
In today's fast moving world, while the gurus are talking about artificial intelligence, it will still take few decades for machines to completely take over decision-making. The process involves applying one's intelligence in their day-to-day business situations, which also keeps the business gain an edge over its competitors.
Earlier, the decisions were mostly based on gut feeling or experience of senior management, but now with enormous pressures and availability of data at speed, business leaders have started relying on data for every decision they take. While the colossal shift in thinking may tire out some people in the beginning, but in the long run, data driven mindset can prove to be a key differentiator for them and lead to continued success for their business.
Some cognitive challenges business leaders face while taking key decisions to drive their business objectives include:-
1. Availability bias
Our brain is a connection machine and we love making connections. It feels good to make links and have everything fit into our mental maps. As a result we create a picture of the world with examples which easily come to our mind. If something is repeated often enough, it gets stored at the forefront of our minds. In turn our mind tends to use short cuts, and based on recent experiences of similar situations or recent memory, one tends to have a bias towards the decision making process. This has been tested by Tversky & Kahneman in their K-study examination as well.
As a strategist, you may be biased from recent information and events that establishes your memory with limited context, leading to an emotional decision than a rational one.
To avoid availability bias, make sure you have access to all the information available before taking any decision. This can be possible if you are good at scanning data at speed. Otherwise a data analyst will come handy, who can constantly look through patterns in data and help you examine the situation purely on merit and not on heuristics.
Also, to come to a conclusion, do a thorough landscape analysis with the help of data. For instance, if I am an online retailer and concerned about taking a decision on whether to build a new App for my business I must first get complete information on the current state i.e. what percentage of my customers are coming via mobile vs desktop? What are the top actions users are taking on the current mobile site? Do users spend more time on Apps? Based on all this information at hand you can take a much more informed decision.
2. Forecast Illusion - False Prophets
As business leaders, we love industry researches to substantiate our hypothesis. And while doing the research, we end up using predictions made by experts and consultants which are aligned to our thought process. By doing this we may fall prey to 'Forecast illusion' because more often than not these predictions go wrong. I am sure at one point you would have believed in the following predictions as well: Hillary Clinton will be the first woman President of USA OR 12-12-12 will be Doomsday, as predicted by hundreds of articles, news bulletins and even by the Mayan calendar.
Philip Tetlock, a Canadian-American political science writer, evaluated 28k predictions from 284 self-appointed professionals and the results in terms of accuracy was that experts were marginally better than random forecast generator. Which leads to a question: Can we trust so called industry experts?
In order to overcome this, one must check on the incentives of the expert while also checking, how good is his/her success rate. It is recommended to work closely with a data analyst, who can look into trends, seasonality (caused by cultural drivers or environmental factors), cyclical movements (caused by macro-economic factors) or any irregularity in various business and economic parameters.
3. Outcome bias
We tend to evaluate decisions based on results rather than on the decision process, which may be misleading especially when the outcomes are unfavorable. Rolf Dobelli's monkey story illustrates the outcome bias very clearly.
If the focus is biased towards past outcomes than the available information at hand, the decision would rely on external factors and not on the process. This would result in poor decision making.Outcome bias is a big risk to an organization's success.
Hence, to avoid outcome bias, one must base the decision making process on information (data) and not on gut feel, or do it in haste. To enable everyone to avoid such biases, organisations must build a data-driven culture. This can be carried through by ensuring availability of relevant data to decision makers (data access), having a single version of truth (data integrity) & educating decision makers about data usage (data literacy).
4. Association Bias
Rick Hanson, in his book 'Hardwiring Happiness' mentioned that our brain hard wires everything it can, which builds to automatic perception. The brain is constantly trying to automate processes, thereby dispelling them from consciousness; in this way, its work will be completed faster, more effectively and at a lower metabolic level. Consciousness, on the other hand, is slow, subject to error and "expensive". This leads to build associations in our brain, which can bias our decision-making ability at speed.
The concept of association is also explained brilliantly by Russian scientist and Nobel prize winner Ivan Petrovich Pavlov. He established the concept of classical conditioning, which is used by all the marketer's around the world in building associations of their brand with human behaviour. His famous experiment on a dog which had learned an association between the sounds of a bell followed by food led to understanding of the association of stimulus to one's thought process.
This kind of bias results in us losing perspective into the actual reasons behind change. While evaluating the reasons leading to an outcome, we must do a thorough attribution analysis of all the factors which may have impacted the results. Attributing the impact to actual reasons would always give a clear view of what has happened. And this in turn would result in better decision making.
One must work closely with a data analyst, who can look through seasonality factors and build attribution models (linear, time-decay, position based, Markov-chain) which would feed into the decision making process.
5. Story Bias
In life, all of our dreams, thoughts & actions are a clutter of details, which we try to fit into a story - a story that follows a pattern and is consistent. We do the same with things happening around us. We try to find patterns and knit it all into a story, which seems rational and uniform. While stories may simplify reality, they may distort it as well by filtering things that do not fit. As decision makers, we need to uncover the truth behind the cause of an event. We should focus not only on factual information, but try and put meaning into it as well. Always rely on data backed stories. Stories may give us a false sense of understanding leading to riskier decisions being made. Various analysis techniques like decision trees, random forest, association analysis, etc. would be a good starting point to understand variance in your key parameters.
6. Fallacy of the single cause
The fallacy of the single cause, also known as causal oversimplification, occurs when we are analyzing an outcome and assuming there is a single simple cause behind it. Interestingly, correlation does not imply causation, which most of the decision makers get trapped in. This can lead to jumping to conclusions quickly and missing the bigger picture of identifying other logical causes. The only way to avoid this cognitive distortion is by using logic to prove a conclusion based on facts backed by data.
Regression analysis and multiple correlation are some of the statistical methods one can apply to come up with actual reasons behind the outcome. Having an analyst who ensures right metrics are being studied, and is able to build the right relationship (patterns) between various factors and can validate the conclusion with data would help you overcome this bias and reach the right conclusions.
7. Introspection illusion
"Introspection illusion," a term coined by Princeton psychologist Emily Pronin to describe the cognitive bias that you don't always know your own thoughts and feelings as well as you think you do.
Now think of business situations wherein the decision maker prefers taking decisions based on introspection. The biggest challenge in such a case is unawareness of error in decision-making. The termed coined by researches for it is 'Choice Blindness' i.e. people lacking insights into their own decision-making process and failing to detect an error. Without effective insights, we can deduce that we have no way to know what & why a decision was made. Typically, we become aware of the decision's fairness only after it is executed, i.e. we can only infer an explanation of it, which may not be an effective way of operating a business in a highly competitive environment.
An analyst supporting the decision making with insights can bring in the validity we seek during the decision making process and being informed about unconscious biases such as introspection bias will lead to a correction effect in the entire process.
As they say that indecision is fatal, even taking a bad decision is as fatal in the competitive world we live in. It is recommended for all the business leaders to create their own checklist to ensure they do not fall for any of these cognitive challenges. These questions would ensure they are at check with reality and governed by precision decision-making backed by data.
Disclaimer: The views expressed in the article above are those of the authors' and do not necessarily represent or reflect the views of this publishing house. Unless otherwise noted, the author is writing in his/her personal capacity. They are not intended and should not be thought to represent official ideas, attitudes, or policies of any agency or institution.