Strategy is at the heart of management and decision-making is at the heart of strategy. So you’d expect most senior leaders to be pretty good at it. However good managers, even great ones, can make spectacularly bad choices, more often than you’d expect. How is that possible, with all the data, decision tools, processes and other matrices available? Bad luck? Bad timing? Poor information?
A large body of research suggests that many are caused by cognitive and behavioural biases.
The biases will happen at 3 levels:
1 – Data selection: the sample is not representative or not all options are available. Good data should give information about both the likelihood of a given outcome and about the value or utility placed on it.
2 – Data presentation: the information is presented in a misleading way, intentionally or not, favouring one outcome over another.
3 – Data interpretation: biases distort our understanding and analysis.
In this article, we’ll cover some of the most common biases and then give recommendations to make better decisions. We will not cover deceptive practices driven by self-protection or conflict of interest, and only cover unconscious biases, the most frequent causes for strategic decisions going wrong.
As Scott Cook, founder of Intuit, puts it, most decisions are based on “politics, persuasion and PowerPoint” – certainly not the most scientific approach!
Rolf Dobelli in his best-selling book “The art of thinking clearly” lists 99 different biases! More business-minded lists are available from the iOpener Institute hereor McKinsey there.
In my experience, though, just a handful are involved in most cases:
Overconfidence and false consensus bias: we tend to think we are better than we really are at making decision. People tend to think that their beliefs, opinions and actions enjoy greater consensus than is really the case.
Overoptimism: high expectations of success, under-estimating the risks or probabilities of the negative scenarii and leading to over-commitment to the chosen option.
Confirmation bias, aka “tunnel vision”: we tend to (unconsciously) focus on information that confirms our view, dismissing (or missing) information that goes the opposite way.
Recency bias: we tend to give more weight to recent data than older data.
Groupthink (and to a larger extent, herd mentality): social pressure to conform means we might tone down our viewpoint if to align to the majority – better avoid “going around the table” or public votes (hands raised).
Champion bias: Aligning to the position of a trusted associate.
Sunflower bias: the boss is always right…!
Survivorship bias: we only take the data we see rather than the data we need. The name comes from WW2 when decision on where to reinforce military airplanes to avoid being shot down was taken on the basis of the planes which came back… rather than those which didn’t!
Anchoring and priming: the starting point becomes the new reference point, changing our reference frame.
Endowment effect and loss aversion: we tend to value more what we already have, and experience losses more acutely than gain. We react differently when the same choice is presented as a loss or as a gain. We tend to avoid risks when negatively framed.
Most big, bad decisions can be attributed to a mix of these biases from the part of the decision-maker, leading to bankruptcies, failed M&As, poor capital allocation, price setting, hiring, or product launches that didn’t live up to expectations. The culprit is usually seen as “not enough information” or “not the right information”, but only a handful of leaders manage to challenge the process itself, to make sure they make the best decision with the information they have.
For an interesting case study, read “A case study in combating bias” in the McKinsey Quarterly dated May 2017, about how German electric utility company RWE overhauled its decision-making processes.
To correct these biases, it is important to rebalance both the input to include more data that contradicts our view and the process with which we make decisions, improving our judgement of the odds, risks and rewards.
Base rate: don’t assume; check the facts and get benchmarks and reference points. Get away from the inside view and take the outside view, where you start with similar cases before considering the specifics of your individual case
Devil’s advocate: appointing someone who has no personal stake in the decision and is senior enough in the organisation to be as independent as possible, to make the negative case
Pre-mortem: imagine the decision was wrong and the project was a failure – what would be the causes and consequences of that failure. Can you do anything to mitigate them? You can expand this for significant decisions into full-blown scenario-planning or war games, putting yourself in competitors’ shoes.
The boss speaks last: get the most junior people to give their opinion first.
The sound of silence: instead of asking people to tell their opinion, ask them to write it down on post-it notes. When they share, it is more likely to remain as per their original thought.
Separate proposals from proposers: get someone else to present and advocate for the proposal.
Select/reject: don’t just try and answer the question “which option should we select?”, but also the opposite one: “which option should we reject?”. People often use different criteria and you might be surprised at the insight
Test and learn: for small, frequent decisions, set up a feedback loop to so as to improve your benchmark
Next-best option: demand alternatives rather than a single yes/no choice.
De-biasing is hard, because it goes against our gut-feel. It takes a lot of self-awareness and discipline to do it right. It’s something you have to constantly force yourself to practice again and again.
Start the process by zooming in on your core decisions, before extending to the corporate culture and ways of working if needed. Where could you decision-making be better? Recruitment? Asset allocation? Look at the numbers: for instance, if you always recruit the same profile, it’s not just because of the market, it might also be due to biases in the process…
For high-frequency decisions, where lots of data is available, AI and analytics can greatly help. But for infrequent yet important decisions, data is often lacking and requires a de-biasing strategy.
As a conclusion, decision-making is more than having proper decision processes and tools in place; it’s about creating an inclusive culture in which people feel safe to speak up and give their opinion, become comfortable with conflict, and feel able to challenge their boss as well as the status quo, so that everybody can be a champion for the right decisions.
Want to learn more?
Get a taste of your full potential: Core Fusion Leadership Development offers a complimentary initial discovery consultation. Sign up for your session today.
Visit our website: www.corefusionleadership.com
Co-authored by Caroline Bonpain and Max Bonpain
Harvard Business Review: Deciding How to Decide, Nov 2013