Want To Improve Any System? Look For It’s Narrowest Point

Want to improve any system? Look for its narrowest point. This is what Rory Sutherland calls Bottleneck Theory. Where does everything slow down? When do talks stall? What’s the decision or who is the decision-maker that can bring everything to a halt? This is the bottleneck. Improvements don’t start until we know where the bottleneck is.


When we have opportunities on one end that we need to transport to the other side, we want to open these points up. This can be difficult to accomplish. Sometimes people want bottlenecks in place for a reason, sometimes for good reasons, other times for not so good ones. These obstacles can be functional, emotional, or social.* Once we know why they’re in place, we can craft a strategy for creating value by opening them up.


Here are some examples of each type:


Functional bottlenecks include capacity constraints and speed bumps. We might add slack back into a system or create a better way for one party to go fast and others to more carefully stay out of the way. Functional bottlenecks have system-design related solutions.


Emotional bottlenecks typically include the ego of a person or persons. We might have to find a way to make them comfortable with loosening their control in an effort to make more progress. Beware illogical and irrational (aka human) behavior here. Emotional bottlenecks have psychological solutions.


Social bottlenecks typically include cultural beliefs and behaviors. We might have to find a way to shift the attitude of an entire group to work differently. Social bottlenecks have cultural solutions. It helps to remember that culture is not a collection of beliefs, but a set of actions. If we want to change them, we’ll need to think carefully about what the group socially values and how to align the desired actions with those values.


Change can come. Find the bottlenecks, ask whether they’re functional, emotional or social, and start working out a strategy. We’ll often find a mix of types, but by labeling each we can consider how to solve for all of the angles. Opening up a bottleneck to create more value is worth the effort.

*not a part of Rory Sutherland’s theory, but I think he’d approve. This is derived from Clayton Christensen’s Jobs Theory detailed in “Competing Against Luck.”