Note: The following are excerpts from A World Without Email by Cal Newport. Passages marked 'MV' are my comments for context, clarity, or readability.
During the year in which I’m writing the bulk of this manuscript, I’m taking a turn as the director of graduate studies for the computer science department at Georgetown. One of the responsibilities of this role is to head the graduate committee, which oversees our graduate program, including approving changes and answering requests about policy. As you might imagine, this leads to a large number of incoming issues that I’m in charge of resolving. Taking a page out of Devesh’s playbook, I internally deploy a Trello board to help make sense of these requests. My board has the following columns:
☐ waiting to deal with
☐ waiting to deal with (time sensitive)
☐ to discuss at next graduate committee meeting
☐ to discuss at next meeting with department chair
☐ waiting to hear back from someone
☐ working on this week
When someone sends me an email or stops by my office with an issue concerning the graduate program, I immediately transform it into a card and place it in the applicable column on the Trello board.
At the beginning of each week, I review this board and move cards around as appropriate: deciding, for example, what I want to work on this week, or what needs to be discussed in upcoming meetings. I can also follow up on issues I’m waiting to hear back from someone about. My general rule is that when I move a card to a new column, I send an email update to the person who brought me that issue. For example, if I move something from a waiting column to the discuss at next graduate committee meeting column, I’ll send a note to the appropriate person telling them that we’ll be discussing their issue soon. If I take a card off the board because I completed the corresponding task, I’ll let the relevant people know the ultimate resolution. And so on.
I have doing and done columns. Following the lead of the Personal Kanban community, I also deploy my own custom blend of columns for making sense of the tasks I plan to work on but am not actively tackling at the moment. Every Monday, I review the board, updating the positions of cards and deciding what I’m working on that week. Throughout the days that follow, I reference the board to figure out what I should do with any time put aside for my DGS duties. As new tasks arrive—in the form of emails, or phone calls, or, as is also common, students dropping by my office to ask questions I don’t know how to answer—I immediately put them on cards that I drop on my board to be dealt with later.
Many proponents of the Personal Kanban approach deploy a single board to make sense of all the tasks in their professional life. I recommend something slightly different: maintain a separate board for every major role in your professional life. At the moment, I play three largely distinct roles as a professor at my university: researcher, teacher, and DGS. I deploy a different task board for each of these roles, so when, for example, I’m thinking about teaching, I’m not also confronted with unrelated tasks about research or the graduate program. This reduces network switching and therefore increases the speed with which I’m able to resolve issues. Similarly, I’ve also found it useful to sometimes set up a dedicated task board for large projects (say, any project that might take more than a couple of weeks of effort). Not long ago, for example, I was the general chair of a major academic conference. The demands of this role were so numerous that I found it easier to contain them on their own task board, isolated from other areas of my academic life. Once that project ended, I discarded the board.
There is, of course, a limit to how many boards you can manage before the upkeep becomes too arduous. This is why I think the rule of one board per role, and one board per major project, is probably about right. For most people this means two to four boards that run your life, which works well. If you have ten boards, on the other hand, the cost of switching between them will begin to swamp out the advantages of separating the tasks.
MV: In A World Without Email, Newport uses the phrase 'The Hyperactive Hive Mind' to describe:
A workflow centered around ongoing conversation fueled by unstructured and unscheduled messages delivered through digital communication tools like email and instant messenger services.
MV: Newport then shares numerous studies documenting the effects of constant communication. Here are six of the best:
A 2011 paper appearing in the journal Organization Studies replicated Mark and González’s pioneering work by shadowing a group of fourteen employees in an Australian telecommunications firm. The researchers found that, on average, the employees they followed divided their workday into eighty-eight distinct “episodes,” sixty of which were dedicated to communication. As they summarize: “These data . . . seem to lend support to the notion that knowledge workers experience very fragmented workdays.”
In 2016, in another paper co-authored by Gloria Mark, her team used tracking software to monitor the habits of employees in a research division at a large corporation and found that they checked email, on average, over seventy-seven times per day. Papers measuring the average number of email messages sent and received per day also show a trend toward increasing communication: from fifty emails per day in 2005, to sixty-nine in 2006, to ninety-two by 2011. A recent report by a technology research firm called the Radicati Group projected that in 2019, the year when I started writing this chapter, the average business user would send and receive 126 messages per day.
A report from the summer of 2018 analyzed anonymized behavior data from over fifty thousand active users of the [Rescue Time] tracking software. It reveals that half these users were checking communication applications like email and Slack every six minutes or less. Indeed, the most common average checking time was once every minute, with more than a third of people checking their inbox every three minutes or less. Keep in mind that these averages are likely inflated because they include periods like lunch breaks and one-on-one meetings in which the subjects were presumably away from their computer screens.
To help understand the true scarcity of uninterrupted time, the RescueTime data scientists also calculated the longest interval that each user worked with no inbox checks or instant messaging. For half the users studied, this longest uninterrupted interval was no more than forty minutes, with the most common length clocking in at a meager twenty minutes. More than two thirds of the users never experienced an hour or more of uninterrupted time during the period studied.
As Leroy hypothesizes, when a task is confined to a well-defined block of time and fully completed during this block, it’s easier to move on, mentally speaking, when you’re done. (Unfortunately for our purposes, when switching back and forth from email inboxes or instant messenger channels, we rarely experience well-defined time limits for our tasks or a sense of completion before switching again.) The more the first task remained on the subject’s mind, the worse they did on the subsequent task. “Every time you switch your attention from one task to another, you’re basically asking your brain to switch all of these cognitive resources,” Leroy explained to me when I asked her about this work. “Unfortunately, we aren’t very good at doing this.” She summarizes the current context in which knowledge workers operate as a state of “divided attention,” in which the mind rarely gets closure before switching tasks, creating a muddle of competing activations and inhibitions that all add up to reduce our performance. In other words, Leroy identified a clear answer to the question that titles her paper. Why is it so hard to do our work? Because our brains were never designed to maintain parallel tracks of attention.
In one particularly devious study, published in 2015 in The Journal of Computer-Mediated Communication, researchers figured out how to discreetly assess our psychological response to thwarted digital connection. Subjects were brought into a room to work on word puzzles. They were told that as part of the experiment, the researcher also wanted to test out a wireless blood pressure monitor. After the subject worked on the puzzle for a few minutes, the researcher returned to the room and explained that the subject’s smartphone was creating “interference” with the wireless signal, so they needed to move the phone to a table twelve feet away—still within earshot, but out of reach. After the subject worked on the puzzle for a few more minutes, the researcher covertly called the subject’s phone. At this point, the subject was trying to solve the word puzzle while hearing their phone ringing from across the room, but was prevented from answering it due to a previous warning from the researcher that it was important not to get up “for any reason.” Throughout this entire charade, the wireless monitor was tracking the subject’s physiological state by measuring blood pressure and heart rate, allowing the researchers to observe the effect of the phone separation. The results were predictable. During the period when the phone was ringing across the room, indicators of stress and anxiety jumped higher. Similarly, self-reported stress rose and self-reported pleasantness fell. Performance on the word puzzle also decreased during the period of unanswered ringing.
Rationally speaking, the subjects knew that missing a call was not a crisis, as people miss calls all the time, and they were clearly engaged in something more important in the moment. Indeed, in many cases, the subject’s phone had already been set to Do Not Disturb mode, which the researchers surreptitiously turned off as they moved the phone across the room. This means that the subjects had already planned on missing any calls or messages that arrived during the experiment. But this rational understanding was no match for the underlying evolutionary pressures which ingrain the idea that ignoring a potential connection is really bad! The subjects were bathed in anxiety, even though their rational minds, if asked, would admit that there was nothing going on in that laboratory that was actually worth worrying about.
There are many signals that work on this social channel. As Pentland explains in his book on the topic, Honest Signals: How They Shape Our World, this information is processed largely unconsciously, often using lower-level circuits in our nervous system, which is why it evades our perceived experience. Its impact, however, shouldn’t be underestimated. “These social signals are not just a back channel or complement to our conscious language,” Pentland writes. “They form a separate communication network that powerfully influences our behavior.” One such signal delivered through this unconscious network is called, aptly enough, influence. It describes the degree to which one person can cause another to match their speaking pattern. This information, which is processed in our brain through subcortical structures centered on the tectum, provides a fast and accurate snapshot of power dynamics in a given room. Another such signal is activity, which describes a person’s physical movements during a conversation. Shifting in your seat, leaning forward, demonstrative gesticulating—these behaviors, which are mediated largely through the autonomic nervous system (“an extremely old neural structure”), provide a surprisingly accurate reading of the true intentions of an individual in the interaction.
It’s not just that we’re less clear than we think, but we’re often completely misunderstood. You were sure that you were sending a nice note, while your receiver is equally sure you were delivering a pointed critique. When you build an entire workflow on exactly this type of ambiguous and misunderstood communication—a workflow that bypasses all the rich, non-linguistic social tools that researchers like Alex Pentland documented as being fundamental to successful human interaction—you shouldn’t be surprised that work messaging is making us miserable.
In 2012, a research team led by Gloria Mark published one of my favorite studies on the impact of email. Their experiment was brilliant in its simplicity: they selected thirteen employees at a large scientific research firm and had them stop using email for five workdays. The researchers didn’t make elaborate contingency plans or develop alternative workflows in advance of the experiment: they simply shut down the subjects’ email addresses and sat back to watch what happened.
As [Gloria Mark] explained to me, one of the subjects was a research scientist who needed to spend around two hours each day setting up a laboratory for an experiment. He reported that he was frequently frustrated because his boss had the habit of sending him emails during this preparation period, asking him questions or delegating work. This required the scientist to stop what he was doing to attend to his boss’s wishes—significantly slowing down the lab setup. The reason Mark remembers this scientist’s plight was because during the five days when he was without email, his boss stopped bothering him during his lab setup. What makes this observation remarkable is that the boss’s office was only two doors down the hall. The small amount of extra difficulty required to walk a few steps and poke his head through the door was enough to prevent the boss from handing off extra work to the scientist. “He was thrilled,” Mark remembers.
MV: Throughout the book, Newport shares case studies of businesses that operate without communication tools (email, Slack, etc.) apart from a single pre-scheduled weekly meeting. One such company is Optimize Enterprises, a self-improvement media company with twelve full-time remote employees. The company's founder, Brian Johnson, tells Newport: "We don't email at all. Zero. There will never be an email between a team member and another team member." Optimize achieves this feat by adopting Henry Ford's assembly line innovation and applying it to knowledge workers. 1913 saw Ford revolutionize manufacturing by asking, 'what if instead of moving the workers between stationary cars, the cars moved past stationary workers?' In Optimize's pipeline, each member waits for a project to become available, performs their single task, and moves on to the next project. Newport explains the entire system:
The process starts with a shared spreadsheet. When Johnson comes up with an idea for a lesson, he adds a title and subtitle to the spreadsheet. Each row has a status column, which Johnson sets to “idea,” marking the lesson as still in the earliest stages of development. Once Johnson gets around to writing the lesson, he’ll upload it to a shared directory in the company’s Dropbox account, then add a link to this draft to the spreadsheet row for the lesson. At this point, he’ll change its status to “ready for editing.” Johnson’s editor doesn’t interact directly with Johnson, but instead monitors the spreadsheet. When he sees a lesson is ready to be edited, he downloads it, puts it into the right format, edits it, and then moves it into a postproduction Dropbox folder that holds text that’s ready to go live.
At this point, the editor changes the status of the lesson to “ready for filming.” Johnson has a studio in his house where he films lessons. He has a standing schedule with his film crew that specifies which days each month they come to knock out a chunk of lesson videos. When the crew arrives, there’s no ambiguity about what they’ll be filming: all lessons currently in the “ready for filming” status. After a film day, the crew will upload the raw files to a shared Dropbox directory dedicated to the editing process. The statuses of these lessons are now changed on the spreadsheet to indicate they are ready to be edited. At this point, Optimize’s film editor will download the clips from the dedicated directory, run them through the standard processing to get them ready for release, and then upload them to a shared postproduction folder. The lessons’ statuses are changed to indicate they are ready for release, and a release date is chosen and added to each corresponding row.
The final step is the actual release of the written and video versions of the lessons on their scheduled release dates. Two content management service (CMS) specialists execute this last step. They monitor the spreadsheet to see which lessons are scheduled for which days. They download the content from the postproduction directories and schedule it for release using the CMS platform. When the time comes, the lesson that started as just an idea in Johnson’s mind goes live across the Optimize networks.
MV: According to Newport, all of the case studies' production processes share three characteristics:
I. It’s easy to review who is working on what and how it’s going.
II. Work can unfold without significant amounts of unscheduled communication.
III. There’s a known procedure for updating work assignments as the process progresses.
A good production process, in other words, should minimize both ambiguity about what’s going on and the amount of unscheduled communication required to accomplish this work.