Stories can inspire, persuade, and transform mundane into memorable. These qualities can make storytelling a potent tool for deception. But a story can also act as a sanity check, a way of confirming and even augmenting one’s understanding.

As a leader, you can use stories to frame problems and motivate. What’s your team’s mission? What will happen if you don’t achieve it? What will happen when you do? What are the obstacles you’ve overcome, and what lies ahead? The answers to these questions don’t fit on a calendar or a Gantt chart. But they uniquely stick in your mind and resonate with your heart.

A story can also be a smokescreen. A charming CEO convinces a group of investors to keep shelling money into a doomed startup. A popular news outlet extracts a small detail of a scientific paper to create a compelling, but misleading headline that goes on to affect thousands of people’s health choices.

In this light, storytelling is at best a convenient, appealing approximation. But for that appeal it sacrifices important parts of the truth. It’s a tool you tolerate to make things happen. And in the wrong hands, it can cause real harm.

But we know this about stories. Less obvious is their tactical application: Stories can act as a sanity check, a sort of integration test for the real world.

Imagine a startup CEO approaching the desk of an engineer who’s been putting in long hours to fix a critical software defect. The CEO asks for a quick summary. The engineer, her mind deep in technical details, attempts a filtered-down explanation. Eventually the executive is able to rephrase this still-detailed description as a story: First we launched feature X, which always had this defect. But it didn’t begin to surface until we later introduced feature Y.

“Got it,” the leader replies. “Can we turn off feature Y for customers affected by this issue?”

The engineer hadn’t thought about this because she was so focused on understanding the root cause and solving it. But the idea is immediately actionable. And it changes the priority of the remaining investigation. If she had asked it sooner, the engineer might have saved herself some time and stress.

(Today this interaction might happen over Slack or email. But that wouldn’t be as good of a story, would it?)

As a manager, you don’t have the option of understanding all the details. So you deal in stories to build a mental model of what’s going on. This can be precarious. But it’s usually good enough. And as the preceding example shows, story-making is not just reductive. It’s often constructive. It can reveal important facets of the work.

Is it possible to have a plan that can’t be summarized? A good product that can’t be explained? A solution you can’t draw lessons from? I’d argue not. These are just signs that you don’t yet understand the plan, product, or solution. So it’s useful to periodically make a story. If you can’t, dig deeper and try again.