Streetlight Effect: Finding Answers Where It Is Easy To Find Them

You look for information where it is easiest to find it. This is known as the streetlight effect or drunkard’s search principle. The origin of the name comes from a story about a policeman who sees a drunk man searching for something under a streetlight. The policeman asks him what he lost. He says his keys. After a few minutes the policeman asked if he is sure he lost them there. The drunk says no, and that he lost them in the park. The policeman then asked why he is searching there. The drunk replies, “this is where the light is”.

Like the drunk, you, your team, or even your entire organization has the tendency to look for answers in the easiest of places. This often leads you to convincing results. Results that are far from the truth. Think about how people typically search the web. Take yourself as an example. Do you use multiple search engines — DuckDuckGo, Google, Brave, Bing? How often do you go beyond the first page? See, you limit your surfing. This way, you seemingly get a satisfying answer while saving energy. Ease, of course, does not always mean accuracy.

It’s natural. When you cannot see in the dark, you look under the only source of illumination. Researchers do this as well. It’s one of the main reasons big data studies don’t apply to the real world. This is because most studies take readily available data based on the assumption that it provides insights into reality, which isn’t always the case. This happened in the field of cardiology in the early 1980s. Anti-arrhythmia drugs burst onto the scene. Heart-attack victims with steady heartbeats have the best odds of survival. A medication that could tamp down irregularities would transform the field.

The drugs became the standard care for heart-attack patients. They were smoothing out heartbeats in intensive care units across the States. But in the early 90s, heart doctors realized the drugs were doing something else as well. They killed about 56 thousand heart-attack patients per year. Hearts were indeed beating more regularly on the drugs. But the problem is, their owners were one-third as likely to survive. Cardiologists focused on the immediately measurable improvement in arrhythmias and overlooked the long-term implication of death. Hence the streetlight effect is a common conversation amongst scientists: most researchers spend their careers looking for answers where the light is better rather than where the truth is likely to lie.

Real-world examples of the streetlight effect:

• The Classic — Lost Keys: a person loses their keys. Instead of searching where they lost them, they look under a streetlight because that’s where its bright. The keys remain unfound. Convenience can result in suboptimal search behavior.

• Academic Research: a scientist uses an easily accessible dataset for analysis. Even though the most relevant data requires a more challenging field study. The conclusions drawn are biased and incomplete because they’re based on the data that was simplest to collect.

• Business Analytics: a company focuses on key performance indicators (KPIs) that are easy to measure. Website clicks, user conversions. Instead of diving into deeper customer engagement metrics. Thus they misunderstand their users’ needs and their campaign is a flop. Lazy analytics, lazy strategy.

• Healthcare: a public health official uses readily available hospital admission data to assess how healthy a community is. They neglect harder-to-measure variables like mental well-being and chronic disease prevalence. The resulting policies address only part of the community’s health needs.

How you might use the streetlight effect as a mental model: (1) question your data — ask whether the information you are using is simply the easiest to get, or if it really represents the complete picture; (2) enter the haunted forest — actively seek out harder-to-obtain data that would offer you a more nuanced perspective, ensuring your conclusions aren’t skewed by convenience; (3) have trust issues — before you finalize a decision, consider if you’re missing information because you looked through the avenues where data was readily available; (4) do what matters — recognize that while it’s tempting to use easy data, sometimes investing extra effort to gather more relevant information saves you from the consequences of big losses.