Forrest Pritchard's farm grows grass, and a farm needs hay, so a hay baler seemed as obvious a purchase as a machine can be. Then, unlike most of us, Pritchard kept counting. Fuel. Repairs. Taxes. Storage. Summer days he could have spent raising chickens for cash and manure. Bought-in hay carried nutrients onto his land, and a heavy machine leaving meant less compacted soil. His real measure of success was long-term fertility, so he sold the baler and bought hay.
Newport puts that story at the center of his argument about tools because almost anything can pass a test that only asks whether it has a benefit. The baler had one. A tool earns its place when its full effect on the few things that matter is clearly positive, and my guess is that most of us have never run that test on a single app in our pockets.
Why 'it has a benefit' decides nothing
Shallow work is the logistical, undemanding stuff you can do while distracted, and Newport's worry is that it expands. It is easy to start, easy to display, and tied to other people's urgency, so it spreads until real work survives in fragments. The software company then called 37signals cut its summer week to four days and found output roughly unchanged, because the missing hours had been disproportionately meetings, interruptions and web surfing. A later month free of memos and status meetings produced a new customer-support toolset and a system for understanding product use.
The other engine of expansion is a belief Newport calls an ideology, that anything on the internet is presumed necessary. A New York Times bureau chief whose value lay in sources and difficult facts was pushed to keep up a stream of lightweight posts. A refrigerated logistics firm painted "like us on Facebook" on its trucks. When readers explained why they used Facebook, they named real but minor gains. Entertainment, light contact with old classmates, a writing group. Then they let one small upside settle the case while compulsive checking stayed off the ledger.
Test it against the few things that matter
Pritchard's way of counting can be written down as a test. Name a few goals in work and life. For each, identify the two or three recurring activities that contribute most. Then judge each tool's effect on those activities as substantially positive, negative, or negligible, and keep it only when the positives clearly win.
Newport runs it on Michael Lewis. His goal might be narrative reporting that changes how people see the world; behind it sit patient research with a few sources and purposeful writing. A social feed contributes little to either and may disrupt both. Run it on friendship and the same thing happens. Long conversations, shared meals and showing up when it counts do the work, and an hour spent monitoring acquaintances may do more as lunch with one close friend.
None of that is a universal verdict. A new student looking for events and people, or a service member overseas keeping in touch with family, may find the same platform central. The high bar comes from an old observation about returns. A minority of activities produces most of what matters, and every minor benefit is paid for out of the same pool of attention as the major ones.
Box it up for thirty days
Habit and fear distort the ledger for tools already in use, so Newport borrows from Ryan Nicodemus, who boxed nearly everything he owned as if he were moving, then took out only what he actually needed over the following week. Most stayed boxed.
The digital version is a thirty-day test. Stop using an optional service, without deactivating it or announcing anything. At the end, ask whether the month would have been notably better with it, and whether anyone cared that you were gone. Return only on a clear yes. The silence is part of the method; a farewell post solicits the attention the test is measuring.
Give every minute a job, then cap the shallow
People are bad at knowing where time goes. British adults aged twenty-five to thirty-four estimated they watched roughly fifteen or sixteen hours of television a week. Measurement put it closer to twenty-eight. So Newport blocks time. Divide the day's hours among projects, routine tasks, meetings and buffers, batch the small things, and redraw the remaining blocks when reality changes. The plan's job is to stop whim and the latest request from silently deciding the day.
Some urgent, important-looking tasks are still shallow, so he offers a test. How many months would a smart recent graduate with no specialized background need to do this well? Editing a research paper might take years to learn, so it is deep. A slide deck from quarterly figures might take a couple of months, so it is shallow, and a status meeting made of milestones and posturing falls the same way. Then set a shallow-work budget, a ceiling on the share of the week such tasks may take, ideally agreed with a manager. Newport suggests many roles land around thirty to fifty percent. The conversation matters more than the number. It confronts both sides with the cost of paying a trained person for easily learned work, and sometimes reveals that instant response is the job.
Fix the inbox upstream
Newport's own inbox is the demonstration. Most email advice is about processing faster; he changes what generates messages. A sender filter states what is relevant, what a sender must provide, and when to expect a reply; he warned that most messages to his address would go unanswered, and serious contact continued while vague requests fell away. A process-centric reply spends a little more thought now to prevent many exchanges later. Instead of "When are you free?", offer a place and several precise times, say that the reply confirms it, and name one fallback. The tone can stay warm.
The last practice makes people flinch. Leave a message unanswered when it is too ambiguous to answer reasonably, when the proposal is irrelevant, or when replying would do no meaningful good and silence no meaningful harm. It is a consequence test, and hierarchy overrides it; a vague note from the chief executive still gets an answer. Minor disappointment is the price of attention having edges.
Leave the evening something better
Removing low-value stimulation leaves a vacuum that the easiest feed will refill unless something better is waiting. More than a century ago Arnold Bennett watched office workers treat paid hours as the real day and the rest as a disposable prologue and epilogue, and asked them to claim a second day within the day. Newport keeps several good books going so reading is the easy default after the children are asleep.
The whole loop is goals, the few activities that drive them, the full cost of every tool against them, a removal experiment when fear blurs the answer, a schedule with a ceiling on the shallow, communication redesigned at the source, and leisure filled on purpose. Coordination survives. Relationships survive. What goes is the baler you never needed, and the assumption that every small benefit gets an unlimited claim on a finite mind.
Three things to keep
Keep a tool only if it clearly helps the few activities behind your real goals.
Quit an optional service for thirty days, quietly, and see who notices.
Give every minute a job, cap shallow work, and reply with next steps.