Reviewing my own decision process

Hello good people. Let’s talk decision-making.

My investments in startups have slowly but steadily added over the last couple of years. It started with African startups - out of all places.

Working with various accelerators in the ANZ region, I have spent a significant amount of time thinking about ways to objectify the due diligence process, optimize for upsides and reduce exposure to the downsides.

This meant critically reviewing my own decision-making process.

The question I asked myself: what are the risks I did not see or chose to ignore? What are the upsides I did not understand?

And if so, why?

My biggest learning: I underestimated the impact of bias and heuristics.

I claim that the real challenge is to understand when to use heuristics, when and for how long to fall back on objective, data-driven processes and how to avoid biased assumptions for decision-making and risk-assessment.

But first, let us look into the differences between data-driven, heuristics and bias.

Heuristics and biases

Heuristics - “methods for arriving at satisfactory solutions with modest amounts of computation” (Oppenheimer, 2008)

• Mental shortcuts or rules of thumb used to make judgments and decisions quickly and efficiently.

• Often based on prior knowledge, experience, or intuition to simplify complex problems or make decisions with limited information.

• Applied to save time and cognitive resources.

Examples:

• availability heuristic - judging the likelihood of an event based on how easily examples come to mind;

• representativeness heuristic - judging the probability of an event based on how similar it is to a typical case.

Biases - an inclination of temperament or outlook; a personal and sometimes unreasoned judgment; prejudice. (Merriam-Webster)

• Systematic errors or deviations from rational judgment that can occur when people rely on heuristics or other cognitive processes.

• They are the brain's attempts to simplify information processing, but often lead to inaccurate and suboptimal decisions.

• Influenced by factors such as emotions, social pressures, and individual experiences.

While there are more than 250 biases, most common cognitive examples:

• confirmation bias - seeking information that confirms one's pre-existing beliefs;

• anchoring bias - relying too heavily on the first piece of information encountered;

• Dunning-Kruger effect - overestimating one’s own competencies.

Heuristics can lead to cognitive biases when they fail to account for relevant information.

However, not all heuristics necessarily lead to biases, but we can get caught very easily.

Heuristic decision process

In 2011 Gerd Gigerenzer from the Max Planck Institute looked into heuristic decision-making. His study suggests that:

1. Heuristics can be more accurate than complex strategies, even though they process less information (less-is-more effects).

2. With sufficient experience, people learn to adaptively select heuristics from their "adaptive toolbox."

3. Heuristics are indispensable in organizational decision-making where the conditions for rational models rarely hold [such as startups and early-stage investments?]

The second point is especially important. Because if this is the case, then experience over time matters to make better decisions. But then I have another question, maybe to be discussed another day: how blind does it make us to creative and disruptive changes?

A critical look in the mirror

With all that in mind, I reviewed my investment notes (calls, research, thoughts).

I looked at all larger decisions, irrespective of whether I finally pulled the trigger and executed / invested or not.

I found, that too often I fell back on heuristics and, somewhat consequentially, bias.

It was really surprising.

Not because I ended up making terrible decisions, although sometimes they could have been better.

But because objectively I relied on bias, when at that time I thought I made decisions based on data or at least heuristics.

Sometimes I didn’t even look at data because I was “gut-feel-sure”.

Maybe because I didn’t know what to look for, or because challenging my own perspective would have concluded in a painful realization that I was (completely) wrong.

Process now

The review impacted the process on how I make decisions now.

For every decision, which requires a significant number of resources (money, time, energy - define your own threshold), I try to distance myself from the process.

Either I ask other people for input on my “problem - process - outcome” or try to challenge myself by asking the following questions:

• Where am I working purely off assumptions?

• Have I actively researched counterarguments to my perspective?

• What are biases I might have fallen into and where could they have tinted my decision?

• What are the heuristics I have fallen back to and are they based on real experience or rather a pretentious ego (hello, Dunning-Kruger)?

• What is the most serious risk factor (not scenario, but factor!) and use data-driven decision-making, while for less impactful risks you can go with heuristics.

A matrix helping with decision-process-analysis could look something like this:

Interestingly, during research on risk-assessment and decision-making under bias, I found the following interview of Greg Fishel, a well-known US meteorologist, and climate-change denier turned climate-change supporter. In this snippet of the interview, he talked about how he slowly started to understand his flawed and biased thinking. It happens to the best of us.

Enjoy the rest of your Thursday.

Best,

Alex