In July 2016 a pandemic broke out. Tiny monsters known as Pokémon suddenly appeared all over the world, threatening to use their fantastic powers to do battle in parks, on city blocks, and in homes. Fortunately, a dedicated volunteer force quickly arose to subdue them, using little-known technology embedded in smartphones to capture and domesticate the creatures.

Pokémon Go was the first big success in what will most likely be a potent run of new multiplayer smartphone games that use augmented-reality technology, which overlays digital images on real-world environments. It was also what we have described as a “big-bang disrupter,” a new product that reigns largely uncontested for a period of success that is often shorter than one would expect from traditional market dominators. (See “Big-Bang Disruption,” HBR, March 2013.)

In the case of Pokémon Go, that period was only a few months. In its first week 7.5 million players downloaded the game. At its peak, just one week later, 28.5 million played for an average of 1.25 hours a day. But 10 weeks later the game had largely run its course. Pokémon Go lost 15 million players in just a month.

By the end of the summer the beasts were gone, along with about $6.7 billion in value for Nintendo, which co-owns the characters that were licensed to Niantic, the developer. Imagining that the $35 million of revenue from players in the first month would continue, investors added $23 billion to Nintendo’s market capitalization, which fell back to earth by August.

Pokémon Go is hardly the only phenomenon that simply ended. Big-bang disrupters such as Fitbit, GoPro, Zenefits, and TiVo scaled up incredibly quickly and then cooled off almost as fast. That’s because they weren’t ready with their next innovation. Companies in that situation may be caught not only with nothing to recapture shrinking revenue but also with their resources committed to a now-faded product. Rapid and total collapse is often the result.

This dramatic rise and even more dramatic fall reminds us of F. Scott Fitzgerald, who famously wrote, “There are no second acts in American lives.” Fitzgerald was referring to the stunning brevity of success in the booming movie industry of the early 20th century, but the same observation is even more applicable to many of today’s hottest businesses.

The good news is that the early demise of so many young enterprises makes it easier to study the underlying causes of modern business failure and their remedies. Using a database of more than 300 big-bang disrupters across multiple industries, we have uncovered important lessons about how to achieve a successful second act.

Though our focus here is on the moment of crisis for start-ups, failure to raise the curtain on a second act isn’t a problem for big-bang disrupters alone. Even the most respected and successful companies in the world today rarely survive their first crisis, whenever it arrives. The average life span of companies on the Standard & Poor’s 500 has fallen from 67 years in the 1920s to just 15 years today. According to Richard Foster, an executive in residence at the Yale Entrepreneurial Institute, in 2020 as many as three-quarters of companies in the index will be companies that were unheard of in 2010.

This shortened life cycle is primarily the result of rapidly spreading digital disruption in industries largely untouched by the first wave of internet transformation—including manufacturing (disrupted by 3-D printing and the internet of things), agriculture (drones and sensors), transportation (autonomous vehicles), and professional services (artificial intelligence). Even if second-act crises are most acute among start-ups, incumbents would do well to understand why they occur and how to avoid them.