This technique was originally developed by
Annie Duke
Read the original content

The Speed Test

Speed is a major determining factor in the decision making process. Unfortunately, both sides of the spectrum carry their own set of risks.

If we move too fast, and we don’t get enough information to cover our blind spots, we risk falling victim to a factor we should have known. On the other hand, if we move too slow, we can get paralyzed in the pursuit of perfect information and incur opportunity costs.

To make matters worse, we have forces working against us when evaluating how quickly to make a decision.

“On an important decision one rarely has 100% of the information needed for a good decision no matter how much one spends or how long one waits. And, if one waits too long, he has a different problem and has to start all over. This is the terrible dilemma of the hesitant decision maker.” Robert K. Greenleaf

Why do we move fast when we should move slow?

Heuristics: Mental shortcuts are hardwired and built over time that create the illusion that the correct choice is clear when we either don’t have enough information or worse — the information we believe we have is biased.

Action Bias: We tend to bias action over inaction. This may cause us to move forward with decisions even when we should wait and collect more information.

Dunning-Kruger Effect: We often believe we know more about a domain that we have very little knowledge or expertise in. In these scenarios, we may not see our blind spots and move forward believing we have a better understanding than we actually do.

Why do we get paralyzed by decisions when we shouldn’t?

Paralysis of choice: The more options we have, the harder it is to make a decision — even with decisions that are seemingly trivial.

Penalties and Payoffs: Our fears around negative outcomes and regret tend to overshadow upside — by a factor of 2x. Understanding our bias to focus on risk over gains can help move forward with accepting and mitigating risk instead of allowing it to delay a decision.

Bikeshedding: We tend to leave complicated or provocative decisions on the shelf regardless of their urgency or importance.

This is by no means a complete and exhaustive list of forces playing against us in the decision making process, but they help to illustrate how some of the forces push us to move slower and others faster.

Using the ‘Speed Test’

The speed test is a tool that helps us quickly isolate decisions that we can make quickly versus decisions that need time — even if we think the decision is clear.

1) Is there any information that can change our mind?

The speed test outlines a very simple question that can effectively segment our decisions,

“Is there information that would change our mind?”.

Poking holes in our decision-making, often referred to a ‘counter-thinking’, is a powerful way to combat bias and check whether or not we’re making a decision based on confidence in the information we have or solely on intuition.

Another way to frame this question is to ask others — especially skeptics — what information would make them more confident. John List, author of the Voltage Effect, recommends recruiting skeptics and objectors to build ‘tests’ that help surface and answer these questions. This method is often referred to as ‘Red Teaming’. [link]

Ask a group of skeptics or detractors what information would change their mind or make them more confident in the decision — and what unknown information is worth moving forward without.

Disagreeableness is a feature in the decision making process, not a bug.

If there is information that would make us, or other contributors, more confident and the time/resources needed is worth a significant boost in confidence, then invest in getting more information.

Adapted from Annie Duke’s book, How to Decide

Move fast if: There is no information that will change your mind (even if you knew that information) or you the information is too costly to attain

Move slow if: What you believe to be true about your assumptions could potentially be false, explore how you could gain confidence.

Don’t make the mistake of chasing perfect information — if the confidence that the decision will provide the expected outcome rises above 70%, move forward.

2) One and two-way door decisions

The concept of ‘one-way door’ and ‘two-way door’ decisions was made popular by Amazon and describes whether or not a decision can be reverted. Just as it sounds, a one-way door decision is incredibly costly to backtrack (if backtracking is even an option) while a two-way door decision has a fallback — or the cost to revert is inexpensive.

For two-way door decisions, it’s helpful to map the ‘penalty’ for a wrong decision. We can ask “What if we’re wrong?” and use an outcome tree [link] to trace the possible negative outcomes, their probability of occurring, and the estimated cost to revert them.

Keep in mind that these these types of decisions are contextual — a one-way door decision for a single team (like choosing a new hire) is often a two-way door decision for the organization that team is in. A new team member has massive implications on the team, but is far less impactful for the company as a whole.

Move fast if: A two-way door decision has low ‘penalty’ for being wrong. Utilize outcome trees to measure the cost and probability of potentially negative outcomes.

Move slow if: Dealing with a one-way door decision

3) Decision boosters

Intuition (but beware)

Decision boosters are tools to help increase decision velocity. One of these tools is simply your intuition. Daniel Kahneman found that some of the highest velocity decision-making processing relied on intuition to make quick decisions, but with a very important caveat; this only works with informed intuition at the end, not the beginning of the process.

Kahneman warns, contrary to popular belief, that intuition only works in environments of extreme repetition and validity.

An executive who has performed countless product launches will likely have better intuition when making go-to-market decisions, but intuition should only help us get across the goal line quicker, not be the basis for the decision.

Repeat decisions

Identify decisions that repeat often and move even faster with them. These are great candidates for rapid experimentation. Examples would be implementing a new shopping cart flow to see a lift in conversions or changing product pricing. These decisions are great candidates for A/B testing or even full automation.

Decision stacking

This is a helpful tool to help reduce risk and increase speed for slow, one-way door decisions. Break the decision down into smaller incremental decisions or ‘validations’ the help reduce risk or gain confidence, then resolve the low-impact decisions first. This pair well with assumption mapping.

Parallel bets

Can bets be made in parallel that allow us to move forward, incur a slightly higher cost, but allow us to quickly get more concrete information? We’re often allergic to ‘throw away’ work, but running parallel experiments is incredibly valuable and low cost relative to doubling down on the wrong strategy.