The printer had nine months. Engineers at the company, which the researcher called “Ditto,” usually took three, four or sometimes five years to develop a product. This one, a color laser printer code-named PEARL and meant to sell for $10,000, was scheduled for nine months, and the division building it was losing money. “If they don’t do it, the whole division will fold,” one manager said.
The researcher Leslie Perlow spent those nine months with the team’s software group, on site an average of four days a week. She shadowed every member and asked the engineers to log their days.
One engineer, whom the paper calls Andy, logged a Wednesday in late November. He arrived at 8:00. At 8:20 he interrupted himself to tell two colleagues they were working on the wrong code. At 9:00 he went to a meeting. At 9:50 a colleague needed help with a new ethernet card. At 10:30 someone came by to update the group on the release plan and ask each person for status, which by Andy’s note was “of some value to him and virtually no value to me.” At 3:05 the ethernet card came back and took an hour and a half. His longest block of solo work came at the end of the day, from 4:50 until 8:15 p.m., when he went home.
Andy’s day was not unusual. Across 35 logs from 12 engineers, three-quarters of their uninterrupted solo stretches lasted an hour or less, and 60 percent of those lasted half an hour or less. Engineers came in early and stayed late looking for quiet, but since most of them did the same, there was always someone around to interrupt.
So Perlow and the engineers tried something. For the first phase, quiet time was set three days a week until noon: no interrupting. A second phase set “interaction time” from 11 a.m. to 3 p.m. every day. A third phase went back to the first design.
Asked to rate their own productivity, 59 percent of the engineers put it above average in phase one, 41 percent in phase two and 65 percent in phase three. That last figure has since been retold, in a 2023 Inc. headline among other places, as a way to increase productivity by 65 percent. It was the share of engineers rating themselves above average.
PEARL launched on time. Three months after the last phase, the division’s vice president credited the quiet-time study. It was the second on-time launch in the division’s history. “I do not think we could have made the deadline without this project,” the vice president said. “This is a new benchmark.”
Perlow kept visiting. Six months after the launch, she found that the lasting effects were minimal. A year later, the engineers told her their next product had slipped behind schedule, and managers had tried quiet time again. One engineer voiced a sentiment Perlow heard from many: “Quiet time alone doesn’t work…. We had no incentive to abide by quiet time…. No one did.”
Perlow traced the relapse to what she called the vicious work-time cycle. Engineers were rewarded for individual heroics in a constant stream of crises, and solving one crisis often meant interrupting someone else, which fed the next crisis. Quiet time had changed the clock but not the rewards.
One thing did stick. Two years after the launch, the managers still checked up on their engineers less often than before.