Confirmation bias is hard to see

When confirmation bias locks you on to a dangerous path

”We don’t believe the world we see; we see the world, we believe”

The Matrix

Monday, September 23, 2013 CEO Thorsten Heins faced one of the biggest decisions in his professional life: Should he end several years of financial struggles and sell Blackberry to the Canadian holding company Fairfax Financial? A special task force had examined strategic alternatives and the best course of action seemed to be to sell BlackBerry at a price of USD 4.7 billion or about 3 percent more than the closing share price Friday. If the deal was accepted, Blackberry would be come a private company away from Wall Street pressure, but four days before the completion of the due diligence Heins was fired and the new CEO, John S Chen, instead raised USD 1 billion cash injection. For a company that only a few years before was the worlds leading smartphone company with 41% of the US market, the highest company valuation in Canada and named the fastest growing company in the world, how could this happen?

BlackBerry’s decline is of course a perfect case study in what happens when a technology giant fails to innovate in a market that is evolving with breathtaking speed. Amid the success investors were warning about the increasing competition from iOS and Android, but BlackBerry maintained strategy. It was only when the iPhone in 2007 began to gain popularity and challenge the BlackBerry that reality hit. But challenge is not one exclusively reserved for fast paced technology markets – in fact other markets may be even more prone to it: When you put a frog into hot water it immediately jumps out, but when you submerge it into normal temperature water, it will stay even when you slowly heat it up (actually, this is not scientifically proven, but it is a great analogy).

The challenge is called confirmation bias: We believe what we see – and we see what we believe. We seek information that confirms our expectations and play down the aspects that are inconsistent with our expectations. It works against innovation, because creative thinkers use information to re-evaluate ideas and avoid status-quo scenarios. Innovative thinking is costly and difficult because confirmation bias is part of our inner mechanics; it is all too easy for us to stop innovate and instead reproduce earlier successful ideas. This is both the reason that giants fall and 90% of new ventures fail – we are only human after all.

Biased information retrieval

We test our hypotheses by searching for confirmatory information consistent with our limited attention and cognitive processing ability that compels us to seek information selectively. This is also one of the reasons that newspaper readership is often split by political leaning – voters prefer to confirm their positions instead of undermining them.

Confirmation bias not only applies to long held beliefs. Often it can be related to an attitude only just developed. For example if we are about to meet someone for the first time and just before the meeting we hear a colleague describe the person as dishonest, then we will immediately start looking for confirming information – and surprise, surprise we find it.

Biased interpretation

Biased interpretation occurs when two people with exactly the same information make different conclusions consistent with previous beliefs. A 1979 Stanford test asked students to evaluate the US death penalty based on research showing the ineffectiveness of capital punishment and both prior proponents AND opponents maintained their previous positions – in fact they left the experiment even more convinced.

Biased memory

Even if we collect and interpret information in a neutral way, we tend to remember it in a way that strengthens our beliefs. Just to consider a hypothesis means that the information stored in our memory consistent with the hypothesis becomes more accessible. Although we often have the feeling that we remember past events correctly – especially when it comes to highly emotional events such as when we first held our baby in our arms, the day we were married or when we were told that we had been promoted – tons of research shows we cannot give an accurate description of the events.

6S Model

To minimize confirmation bias it critical that your organization is geared to be open and even proactive towards alternatives, surprises and disagreements across the 6S parameters of Strategy, Structure, Steps, Systems, Skills and Style.

In Strategy there is a myriad of tools, concepts and even schools, but not all are equally valuable or relevant to your business. It is not always obvious what is the strongest solution, but once you have bought into a particular tool, you begin automatically searching for affirmative arguments that exactly the strategy method you have chosen is the best hammer for all your different challenges. Playing to Win is a strong approach, because each step from defining winning ambition over selecting where to play and how to win to designing core capablities and management system, are designed to test the previous step.

In Steps (aka processes) we face repeated decisions, where the answer is often the same. This reinforces the confirmation bias – it takes a lot to answer no when you just answered yes 1,000 times. The classic scientific and consultant problem solving approach is to first look for how you can disprove your hypothesis. A simple way to ensure this are checklists spurring just enough conscious thought to avoid disasters.

In Style (aka culture) we cannot stress the importance of diversity enough, but there will be no positive results unless you insist on maintaining the constructive disagreement arising from diversity. Even with a healthy diversity approach, individuals may unconsciously sabotage the opportunity, e.g. when the manager starts out with his or her opinion and then asks for alternative views – funnily enough you do not get a good discussion going. In this big data focused age, it is important to remember that while facts are critical then organizations only focused on data are paving the way their only personal confirmation bias hell. Data can not replace common sense and good decision processes -not everything that can be measured is important and not everything that is important can be measured.

This was the abbreviated chapter 4 of our acclaimed book “Decision Strategy” – next two weeks we take a summer break. Cant wait? Then contact brian@behaviouralstrategygroup.com or +45-23103206.