The phrase “bias towards action” is popular in knowledge work. It means do something before overanalyzing what you should do.

I’m not sure who officially coined the phrase, but Google and Amazon made it popular. From Amazon’s values:

“Bias for Action: Speed matters in business. Many decisions and actions are reversible and do not need extensive study. We value calculated risk-taking.”

A bias for action is useful. Especially in organizations that have a tendency to plan and are resistant to trying new things. I’ve seen my previous clients try new initiatives and be more open-minded because of adopting this principle.

However, I’ve also seen the danger of taking a “bias toward action” to the extreme. This essay will share three pitfalls of extreme bias toward action and suggest that sensing before acting is more helpful than biasing towards action.

Bias Toward Action Traps

Trap #1: Solutions that don’t solve actual problems.

Take this example. Jane, a designer, shares an issue with her team in their weekly retrospective. Her project is drastically behind schedule. Ellen, a more senior colleague, immediately suggests to Jane that she should draft a Gantt Chart as a way to get back on schedule. Ellen wants to help Jane, biases toward action, and loves Gantt Charts. Jane, out of respect for her more senior colleague, agrees to draft the Gantt Chart. Discussion over.

However, drafting a Gantt Chart doesn’t solve Jane’s problem. The real problem is that the client has not decided which of their products will be available for launch, blocking the design of Jane’s website. The real solution isn’t a Gantt Chart. It’s having a real conversation with their client, explaining how their project is blocked until they know which exact products are available for launch.

Ellen’s “bias towards action” prompted a solution that didn’t solve Jane’s problem. Creating the Gantt Chart would be a waste of her time.

Trap #2: No time for strategy.

A large organization I consulted with used to end every meeting with clear actions and owners assigned to each action. If the meeting didn’t have actions and owners, it wasn’t a successful meeting. The reason? “If we’re moving slow, we’re losing.” No one had a clear answer when someone asked what larger strategy the actions served. Strategy was a nice-to-have in this organization (they realized this and have since prioritized doing more strategy work).

A bias toward action can perpetuate the perception that there’s no time for higher-level thinking. We always need to be doing something to move the metric, get our feature out before our competitor, or fully utilize our resources. It can lead to working towards outputs instead of outcomes. If we don’t do any strategy, we’re going nowhere, fast.

Trap #3: We make irreversible decisions without crucial input.

The most dangerous bias towards action trap is making an irreversible decision without crucial input. Layoffs. An acquisition. A company-wide reorg.

These decisions are often made without input from folks who have context and expertise and/or will be most impacted by the decision. Why? You guessed it, a bias towards action.

To be fair, most organizations benefit from making reversible decisions quickly. The keyword is reversible. Jeff Bezos suggests that reversible decisions should be made quickly, while irreversible decisions should be made more deliberately. From Amazon’s 1997 letter to shareholders:

“Some decisions are consequential and irreversible or nearly irreversible — one-way doors — and these decisions must be made methodically, carefully, slowly, with great deliberation and consultation. If you walk through and don’t like what you see on the other side, you can’t get back to where you were before. We can call these Type 1 decisions. But most decisions aren’t like that — they are changeable, reversible — they’re two-way doors. If you’ve made a suboptimal Type 2 decision, you don’t have to live with the consequences for that long. You can reopen the door and go back through. Type 2 decisions can and should be made quickly by high judgment individuals or small groups.

As organizations get larger, there seems to be a tendency to use the heavy-weight Type 1 decision-making process on most decisions, including many Type 2 decisions. The end result of this is slowness, unthoughtful risk aversion, failure to experiment sufficiently, and consequently diminished invention.1 We’ll have to figure out how to fight that tendency.”

Of course, Jeff is saying that we should default to action for Type 2 (reversible) decisions. But again, it’s important to distinguish reversible from irreversible decisions. Another way of putting it is W.L. Gore’s Water Line Principle: decisions that are “above the waterline” won’t sink our ship, while decisions “below the waterline” will. Above the waterline decisions should be made quickly and autonomously. Below the water line decisions should be made thoughtfully and with input.

From W.L. Gore’s principles:

“Waterline: We are all shareholders, and we will consult with the appropriate Associates before taking an action “below the waterline” that could cause serious damage to the long-term success or reputation of our Enterprise.”

The right action depends on the problem type

So when is it beneficial to bias towards action? And when is it not?

Here’s a helpful way of thinking about it: how and when we act should depend on our context. Let’s use the Cynefin Model to unpack this.

The Cynefin (pronounced kuh-NEV-in) Model, developed by systems theorist Dave Snowden, is a framework that helps you think through a situation and figure out the best response to it. It lays out four domains: Obvious, Complicated, Complex, and Chaotic. (Its fifth domain, Disorder, is when the problem doesn’t fit any of the four.)

Let’s walk through each domain. From Phil Zofrea:

“Obvious (the known) — We’ve seen this a million times and as such can categorize and respond according to established best practices. The relationship between cause and effect is well known.”

Examples of obvious problems:

You follow an online recipe to make carbonara.
Your computer froze, so you restart it.
Your team agreed to a date and time to meet, so you send a calendar invite.

Because the relationship between cause and effect is well known, best practices obvious simple problems.

“Complicated (the knowable) — Although we don’t immediately know what is happening, we can analyze the situation and come to a conclusion of what must be done. We can enlist experts to analyze, set up constraints, and a process addressing resolution.”

Examples of complicated problems:

You’re fixing a car.
You’re implementing a customer relationship management tool.
You’re analyzing whether there was a statistically significant difference in your a/b test.

Analyzing before acting solves complicated problems.

The difference between complicated and complex problems is that complicated problems have a known solution while complex problems do not.

“Complex (the unknowable) — We’re not able to determine what will cause a particular result. The best course of action is to conduct experiments and check if any or all take us in the correct direction. A lot of time when human opinion and decision is involved we could be working in this area; simply because humans are complex beings.”

Examples of complex problems:

You’re looking to disrupt the automotive industry.
You want your startup to achieve product-market fit.
You’re a product leader at Spotify, trying to increase profit margins by hosting podcasts.
Your team is figuring out the best way to work together as a distributed team.

Small experiments solve complex problems. Because we don’t know what action will solve a complex problem, our goal is to get data quickly on solutions that may work.

“Chaotic (the incoherent) — The situation is very unstable. We don’t have time to experiment or probe since the situation is dire and we need to act. An IT issue that must be taken care of immediately with no delay may be categorized as such. If we have no time to figure out a system deadlock issue, we may opt to get ourselves out of this chaotic state by rebooting the server.”

Chaotic examples:

Your house is on fire and your brother is inside.
Your city is preventing the spread of coronavirus by social distancing or mass testing.
A server that supports the product of a million users is down. You’re resolving this situation.
Your 450-person company has no cash flow and no cash on hand.

In a chaotic situation, there’s no time to analyze. You must act. Snowden and Boone suggest:

“A leader must first act to establish order, then sense where stability is present and from where it is absent, and then respond by working to transform the situation from chaos to complexity, where the identification of emerging patterns can both help prevent future crises and discern new opportunities.”

Covid-19 is a great example of a chaotic problem. To reduce the virus’s spread, we have to socially distance. We have to act now, while we figure out a better solution (ubiquitious screening, mass-produced tests, a vaccine).

Before you act, understand whether your problem is simple, complicated, or complex. If it’s chaotic, just act.

Know your context

Overbiasing towards action is like shooting a basketball without looking at the hoop. You’re shooting without aiming.

A team that over biases toward action can fall into one of three traps. They waste time on solutions that don’t solve actual problems. They don’t take time to strategize. And they make irreversible decisions with no input.

Instead of always biasing towards action, know the context of your problem. Is it simple, complicated, complex, or chaotic? Then act.