I was about ready to abandon this book during the first part, in which Newport describes all the reasons why a workflow focused on email and chat is terribly unproductive. It felt too drawn out, and although I’m certainly in agreement that the constant distraction imposed by electronic messaging isn’t good, some of his arguments were weak.

But I wanted to see what suggestions he’d have for alternative workflows, and I’m glad that I persevered; the productivity suggestions in the second half of the book are useful and broadly applicable. I’ve never really been subjected to a “hyperactive hive mind” messaging-based workflow, but I’m hoping to make some changes based on Newport’s advice anyway.

Book outline

I created an outline of the book’s main points so that I can review them in the future, and I figure that I may as well share it. Text in bold, italics, or quotes is lifted directly from the book; everything else is my own words.

Part 1: The case against email

Newport argues that the “hyperactive hive mind” – near-constant use of email and other electronic messaging tools for excessive, unscheduled communication and coordination – reduces productivity, makes us miserable, and has a mind of its own. I didn’t really enjoy this part and won’t bother with the details.

Part 2: Principles for a world without email

This part is where all the good stuff is. Newport describes four guiding principles for more productivity in knowledge work and suggests ways to implement each.

The attention capital principle

“The productivity of the knowledge sector can be significantly increased if we identify workflows that better optimize the human brain’s ability to sustainably add value to information.”

Newport compares knowledge work with industrial work, noting how industrial work saw huge productivity increases during the twentieth century by establishing processes that made better use of capital. The primary example he uses is Ford’s assembly line.

  • Build structures around autonomy. Knowledge workers need autonomy to do their jobs well, but we should distinguish between autonomy in workflow – how their work is organized – and execution – how their real work gets done. The latter is important; the former is not and should be optimized.

  • Minimize context switches and overload. Industrial work has a clear optimization metric: speed. Knowledge work’s goal is more ambiguous: reduce context switching and the overload that comes from disorganized work.

  • Don’t fear inconvenience. Electronic communication is super convenient, but sometimes that’s a bad thing. A bit of friction can push people to be more independent or develop more efficient processes rather than simply firing off a message whenever a question or problem arises.

  • When implementing changes, seek partners, not forgiveness. Changing workflows affects other people. When your changes affect how others do their work, make sure they know about the changes, are involved in deciding what changes get made, and have a way to improve the workflow in the future. When your changes affect others’ expectations of you (e.g., longer response times), don’t make a big deal out of them; just do good, reliable work and people generally won’t care how you do it.

The process principle

“Introducing smart production processes to knowledge work can dramatically increase performance and make the work much less draining.”

Newport uses the example of a media production company that relies on processes based on spreadsheets and task boards to coordinate their work. He argues that well-designed processes are much better than ad-hoc messaging.

  • Properties of effective processes.

    1. “It’s easy to review who’s working on what and how it’s going.”
    2. “Work can unfold without significant amounts of unscheduled communication.”
    3. “There’s a known procedure for updating work assignments as the process progresses.”

  • Cards in columns. Task boards that use columns of cards (e.g., a Kanban board) are often effective means of organizing and coordinating work, with the following guidelines:

    • “Cards should be clear and informative.”
    • “When in doubt, start with Kanban’s default columns.” These are to do, doing, and done.
    • “Hold regular review meetings.” Meet synchronously with your team on a regular schedule to go through the task board, adding/removing/moving cards as needed.
    • “Use card conversations to replace hive mind chatter.” Some software for task boards (like Trello) allows users to comment on cards. These comments keep communication about a specific project/task in one place rather than scattering it around various email or chat conversations.

  • Personal Kanban: Organizing your professional life with individual task boards. While the previous section focused on using task boards for team work, they can also work well for individual task management. Again, some guidelines:

    • “Use more than one board.” Separate boards for separate roles or large projects can be useful.
    • “Schedule regular solo review meetings.” Have a scheduled time to review your task board(s) just as you would in a team.
    • “Add a ’to discuss’ column.” Use this column to keep track of things you need to talk to others about. Once you have a decent group of things to talk about with someone, schedule a time to do so face-to-face or over video.
    • “Add a ‘waiting to hear back’ column.” Use this column to hold onto tasks that aren’t fully done but aren’t in your court at the moment.

  • Automatic processes. Some work can be coordinated without any messaging at all. The example at the beginning of the chapter describes how the media production company coordinates video production via a spreadsheet; at each phase of the process, files are uploaded to a cloud service and the status of the video is updated in the spreadsheet, allowing another person to begin the next step. Automatic processes should accomplish the following:

    • Partitioning: “Split the process into a series of well-defined phases that follow one after the other.”
    • Signaling: “Put in place a signaling or notification system that tracks the current phase of each output.” This doesn’t have to be invasive notifications – if everyone checks the spreadsheet regularly, for example, simply updating a column in it could be sufficient to signal that it’s time to begin the next phase.
    • Channeling: “Institute clear channels for delivering the relevant resources and information from one phase to the next.” For example, shared folders in a cloud service.

  • Making individual work automatic. Individuals can also “automate” some of their processes by setting aside specific times to work on recurring tasks/activities. This consistency can reduce decision overhead.

The protocol principle

“Designing rules that optimize when and how coordination occurs in the workplace is a pain in the short term but can result in significantly more productive operation in the long term.”

Using examples from information theory and computing, Newport describes how often-repeated tasks deserve smart protocols that keep them efficient. The upfront cost of working out these protocols is outweighed by the time and energy they save in the long term.

  • Meeting scheduling protocols. Use a hired assistant or a service like Calendly to avoid the back-and-forth of scheduling meetings.

  • Office hour protocols. Set aside regular blocks of time when you’re available for one-off questions and conversations rather than allowing those discussions to happen anytime over email/chat.

  • Client protocols. Develop clear expectations with clients about how you’ll communicate with them. Ideally, make it possible for them to get status updates and answers to questions without interrupting you. (For example, a ticketing system that gets updated as work is done can allow clients to login and see the status of their ticket rather than sending an email.)

  • Non-personal email protocols. Use project- or topic-specific email addresses rather than person-specific email addresses to change people’s expectations about responsiveness.

  • Short-message protocols. Intentionally set a limit on how long emails will be in order to push non-trivial discussions, which aren’t very efficient over email, to synchronous meetings.

  • Status meeting protocols. Hold frequent, short status meetings in which everyone reports on what they’ve done since the last meeting, what help they need, and what they’ll do next. These meetings increase accountability and keep everyone on the same page without causing constant interruptions.

The specialization principle

“In the knowledge sector, working on fewer things, but doing each thing with more quality and accountability, can be the foundation for significantly more productivity.”

Newport explains how computers have failed to increase knowledge worker productivity because they’ve made doing one’s own administrative work easier; businesses hire less support staff and instead expect people to spend significant time doing things outside their specialty. Extreme programming is an example of how allowing people to spend nearly all of their time focused on what they do best can dramatically increase productivity.

  • Do less, do better. Outsource the things that aren’t within your area of expertise to others, perhaps by hiring support staff or paying for a service. Individual employees can sometimes get the freedom to be selective about what they work on at the cost of higher-stakes accountability – you won’t look as busy, so you’ll have to do your real work well.

  • Sprint, don’t wander. Focus intensely on a single thing for a while (e.g., a few weeks) rather than jumping back and forth between different projects/tasks. When a whole team does this together, it’s especially effective.

  • Budget attention. Keep track of how long you spend on work that isn’t your specialty, and set limits. Others, including supervisors, are often more willing to let you off the hook for small requests if they realize how much of your time these non-core activities are taking in aggregate.

  • Supercharge support. When everyone is more focused on their specialty, extra support staff is often needed to handle what’s left. The following guidelines help make this extra support effective:

    • “Structure support.” Ensure that support staff don’t suffer from the hyperactive hive mind either by structuring their work well.
    • “Build smart interfaces between support and specialists.” Design support processes primarily to free up specialists’ time and energy, even if that means the processes aren’t as efficient for the support staff.
    • “As a last resort, simulate your own support staff.” As an individual without support staff, set aside specific time blocks to do your own support work. Use the previous principles to make this support time effective.