When: The Scientific Secrets of Perfect Timing

by Daniel H. Pink

What Kahn discovered in the job market is what chaos and complexity theorists have long known: In any dynamic system, the initial conditions have a huge influence over what happens to the inhabitants of that system.28

For example, in 2010 four social scientists, including Nobel Prize–winning economist Angus Deaton, took what they called “a snapshot of the age distribution of well-being in the United States.” The team asked 340,000 interviewees to imagine themselves on a ladder with steps numbered from 0 at the bottom to 10 at the top. If the top step represented their best possible life, and the bottom the worst possible one, what step were they standing on now? (The question was a more artful way of asking, “On a scale of 0 to 10, how happy are you?”) The results, even controlling for income and demographics, were shaped like a shallow U, as you can see in the chart.

Tillery and Fishbach also engaged other participants in what they claimed was a test of how young adults perform on skills they hadn’t used much since childhood. They handed people a stack of five cards, each of which had a shape drawn on it. The shape was always the same, but it was rotated into a different position on each card. They gave people scissors and asked them to cut out the shapes as carefully as possible. Then the researchers presented the cutout shapes to lab workers not involved in the experiment and asked them to rate, on a 1-to-10 scale, the cutting accuracy of the five shapes.

The result? Participants’ scissor skills rose at the beginning and end but slumped in the middle. “In the domain of performance standards, we thus found that participants were more likely to literally cut corners in the middle of the sequence rather than at the beginning and end.”

Since Gersick obtained results she didn’t expect, and since those results ran counter to the prevailing view, she searched for a way to understand them. “The paradigm through which I came to interpret the findings resembles a relatively new concept from the field of natural history that has not heretofore been applied to groups: punctuated equilibrium,” she wrote. Like those trilobites and snails, teams of human beings working together didn’t progress gradually. They experienced extended periods of inertia—interrupted by swift bursts of activity. But in the case of humans, whose time horizons spanned a few months of work rather than millions of years of evolution, equilibrium always had the same punctuation mark: a midpoint.

For example, Gersick studied one group of business students given eleven days to analyze a case and prepare an explanatory paper. The teammates dickered and bickered at first and resisted outside advice. But on day six of their work—the precise midpoint of their project—the issue of timing parachuted into the conversation. “We’re very short on time,” warned one member. Shortly after that comment, the group abandoned its unpromising initial approach and generated a revised strategy that it pursued to the end. At the halfway mark in this team and the others, Gersick wrote, members felt “a new sense of urgency.” Call it the “uh-oh effect.”

She found the same dynamic over longer periods. In other research, she spent a year following a venture-capital-backed start-up company that she called M-Tech. Entire companies don’t have the finite lives or specific deadlines of small project teams. Yet she found that M-Tech “showed many of the same basic temporally regulated punctuational patterns as project groups show, on a more sophisticated, deliberate level.” That is, M-Tech’s CEO scheduled all the company’s key planning and evaluation meetings in July, the midpoint of the calendar year, and used what he learned to redirect M-Tech’s second-half strategy.

Halftimes in sports represent another kind of midpoint—a specific moment in time when activity stops and teams formally reassess and recalibrate. But sports halftimes differ from life, or even project, midpoints on one important dimension: At this midpoint, the trailing team confronts harsh mathematical reality. The other team has more points. That means only matching them in the second half will guarantee a loss. The team that’s behind must now not only outscore its opponent, it must also outscore the opposition by more than the amount it’s trailing. A team ahead at halftime—in any sport—is more likely than its opponent to win the game. This has little to do with the limits of personal motivation and everything to do with the heartlessness of probability.

Jonah Berger of the University of Pennsylvania and Devin Pope of the University of Chicago analyzed more than 18,000 National Basketball Association games over fifteen years, paying special attention to the games’ scores at halftime. It’s not surprising that teams ahead at halftime won more games than teams that were behind. For example, a six-point halftime lead gives a team about an 80 percent probability of winning the game. However, Berger and Pope detected an exception to the rule: Teams that were behind by just one point were more likely to win. Indeed, being down by one at halftime was more advantageous than being up by one. Home teams with a one-point deficit at halftime won more than 58 percent of the time. Indeed, trailing by one point at halftime, weirdly, was equivalent to being ahead by two points.20

Truckloads of sports data can reveal correlations, but they don’t tell us anything definitive about causes. So Berger and Pope conducted some simple experiments to identify the mechanisms at work. They gathered participants and pitted each one against an opponent in another room in a contest to see who would bang out computer keystrokes more quickly. Those who scored higher than their opponents won a cash prize. The game had two short periods separated by a break. And it was during the break that experimenters treated their participants differently. They told some that they were far behind their opponent, some that they were a little behind, some that they were tied, and some that they were a little ahead. The results? Three groups matched their first-half performance, but one did considerably better—the people who believed they were trailing by a little. “[M]erely telling people they were slightly behind an opponent led them to exert more effort,” Berger and Pope write.21

Third, at the midpoint, imagine that you’re behind—but only by a little. That will spark your motivation and maybe help you win a national championship.

For example, to run a marathon, participants must register with race organizers and include their age. Alter and Hershfield found that 9-enders are overrepresented among first-time marathoners by a whopping 48 percent. Across the entire life span, the age at which people were most likely to run their first marathon was twenty-nine. Twenty-nine-year-olds were about twice as likely to run a marathon as twenty-eight-year-olds or thirty-year-olds.

The encoding power of endings shapes many of our opinions and subsequent decisions. For instance, several studies show that we often evaluate the quality of meals, movies, and vacations not by the full experience but by certain moments, especially the end.19 So when we share our evaluations with others—in conversations or in a TripAdvisor review—much of what we’re conveying is our reaction to the conclusion. (Look at Yelp reviews of restaurants, for example, and notice how many of the reviews describe how the meal ended—an unexpected farewell treat, a check with an error, a server chasing after diners to return an item left behind.) Endings also affect more consequential choices.

The researchers gave half their participants the bad-guy-to-good-guy bio and half the good-guy-to-bad-guy bio, and asked both groups to evaluate Jim’s overall moral character. Across multiple versions of the study, people assessed Jim’s morality based largely on how he behaved at the end of his life. Indeed, they evaluated a life with twenty-nine years of treachery and six months of goodness the same as a life with twenty-nine years of goodness and six months of treachery. “[P]eople are willing to override a relatively long period of one kind of behavior with a relatively short period of another kind just because it occurred at the end of one’s life.”22 This “end of life bias,” as the researchers call it, suggests that we believe people’s true selves are revealed at the end—even if their death is unexpected and the bulk of their lives evinced a far different self.

I’ve got some good news and some bad news.” You’ve undoubtedly said that before. Whether you’re a parent, a teacher, a doctor, or a writer trying to explain a missed deadline, you had to deliver information—some of it positive, some of it not—and opened with this two-headed approach. But which piece of information should you introduce first? Should the good news precede the bad? Or should the happy follow the sad? As someone who finds himself delivering mixed news more often than he should or wants to, I’ve always led with the positive. My instinct has been to spread a downy duvet of good feeling to cushion the coming hammerblow. My instinct, alas, has been dead wrong.

We blunder—I blunder—because we fail to understand the final principle of endings: Given a choice, human beings prefer endings that elevate. The science of timing has found—repeatedly—what seems to be an innate preference for happy endings.30 We favor sequences of events that rise rather than fall, that improve rather than deteriorate, that lift us up rather than bring us down. And simply knowing this inclination can help us understand our own behavior and improve our interactions with others.

The crux of the experiment came just before people tasted the fifth chocolate. To half the participants, the research assistant said, “Here is your next chocolate.” But to the other half of the group, she said, “Here is your last chocolate.” The people informed that the fifth chocolate was the last—that the supposed taste test was now ending—reported liking that chocolate much more than the people who knew it was simply next. In fact, people informed that a chocolate was last liked it significantly more than any other chocolate they’d sampled. They chose chocolate number five as their favorite chocolate 64 percent of the time (compared with the “next” group, which chose that chocolate as their favorite 22 percent of the time).

Bonus: If you’ve got an extra minute left, send someone—anyone—a thank-you e-mail. I mentioned in chapter 2 that gratitude is a powerful restorative. It’s an equally powerful form of elevation.

Human beings rarely go it alone. Much of what we do—at work, at school, and at home—we do in concert with other people. Our ability to survive, even to live, depends on our capacity to coordinate with others in and across time. Yes, individual timing—managing our beginnings, midpoints, and endings—is crucial. But group timing is just as important, and what lies at its heart is crucial for us to know.

What Adhav does is fundamentally different from delivering a Domino’s pizza. He sees one member of a family early in the morning, then another later in the day. He helps the former nourish the latter and the latter appreciate the former. Adhav is the connective tissue that keeps families together. That pizza delivery guy might be efficient, but his work is not transcendent. Adhav, though, is efficient because his work is transcendent.