Tag Archives: uncertainty

This could only mean one of 347 things


‘Why’ is a frequently answered question

Attribution is about figuring out the cause of events and answering the question why something happened. When it comes to our social lives, all humans are natural-born attribution-making machines.

Why did I lose my job? → It’s the economic downturn, and the company is laying off people.

Why is my boyfriend breaking up with me? → He is going off to college, AND he found someone else.

Why am not spending enough time with my family? → My boss is keeping me busy at work.

Of course, we tend not to answer the question for those events that have a widely accepted explanation, e.g. the sunrise, and those that don’t affect us personally, e.g. the death of a neighbor’s cat. Everything else is fair game for our innate attribution processes.

The basic function of these judgments is to help us make sense of the world, and we make a staggering number of them every day, often without realizing it. However, what’s really important is that most of the reasoning we dish out is subject to what psychologists call attribution bias. It’s not that we are stymied by less than perfect inputs in making the right judgements–it’s that we are predisposed to making certain kinds of errors even when the inputs are perfect.

What does all of this have to do with startups?

Eric Ries defines a startup as “a human institution designed to deliver a new product or service under conditions of extreme uncertainty”. When I think of this definition I think of uncertainty as something that’s coming at me from the future. Attribution biases mean that the past is chock full of uncertainties as well because our interpretation of the past is laden with erroneous judgments.

Wikipedia’s attribution bias article currently links to 13 types of different attribution biases. Most of them have implications for startups. For starters, consider these 4.

  1. Self-serving bias and its close relative group-serving bias. I only have two words: pervasive and pervasive. Essential this bias means that people attribute their successes to internal or personal factors and their failures — to situational factors beyond their control.
  2. Egocentric bias, aka “I did it all” bias. We tend to diminish the contribution made by others while exaggerating our own. Need I say more?
  3. False consensus effect. Yep, that’s thinking that everybody else thinks just the way you do.
  4. Fundamental attribution error. This bias makes you prefer personality-based explanations to situation-based explanations.

Imagine now the havoc these biases can wreak in an environment of “extreme uncertainty”: the great temptation to attribute the bump in site’s traffic to the inherent value of your offering over the press coverage that gave you a temporary boost, or the temptation to think that if you like an idea, or you solve your own problem when you build something, then others too will find want your solution.

Lean startups and attribution biases

Build, measure, learn virtuous cycle of a lean startup is straightforward enough until you realize that the learn part of the loop (what you learn for the next build phase) is fertile ground for all sorts of attribution biases. Iterations hinge on the fact that you first measure things that matter (see the great post on vanity metrics, also by Eric) and then extract knowledge that feeds into your next cycle through the loop. Subvert the learn phase, and you derail the whole iteration.

All businesses are prone to attribution bias, but only startups rely on continuous learning and small iterations to extract the business knowledge they need to succeed. Recognizing learning as a startup’s primary driving force is one thing, reining in individual biases and allowing learning to take place is another.