The setting takes four seconds to fill in. A time, and a number. Most people type 7:00 and once. Seven is when they get up, and once is what the phrase morning briefing means. That’s a workable answer, and the choice then stands for a year.
It’s a larger decision than the form makes it look. The chapter before this one ended by pointing straight at it. Every line on the exception list you just wrote covers a gap. These two fields set the gap. Move them and half the list can come off. Leave them and the list carries nine working hours on its own.
So bring the list. This is the chapter where the two get tested against each other, and each one settles the other.
- What you’re exposed to is your longest gap. Two reads placed an hour apart leave a longer one than a single read placed well.
- The hour comes off your calendar rather than your alarm clock. Work back from the first fixed thing today, and subtract whatever the briefing might ask you to do about it.
- The window has two edges. Late, and it lands after the moment it was meant to protect. Early, and it describes a night instead of a day, so you open the inbox at 9:20 anyway.
- A second read is paid for in lines retired from the exception list. Count them first. A line that fires twice a month is cheaper to keep than a read you take 250 times a year.
- Above two or three, the arrangement is spent. Name that ceiling in advance and it holds against the drift back to the inbox, one reasonable addition at a time.
One thing to clear out of the way, because two features share a verb. This chapter is about when a briefing reaches you. Scheduling a message of your own to go out at a set hour is a different feature wearing the same word, and this chapter leaves it alone.
Your exposure is one number, and it is not the count
The whole decision reduces to a single quantity: the longest stretch in which something can arrive, matter, and sit unseen. That’s the number to work with, rather than the average stretch or the number of times you read.
Two adjustments make that quantity honest.
Count only the hours that could contain something. An overnight gap of fourteen hours and a daytime gap of four sit in different classes. Mail at 2:00 in the morning is rare, and anything that does land waits for eight o’clock either way. So price a gap by the working hours inside it rather than the clock hours. That’s what makes the overnight stretch mostly free, and it’s the reason a once-a-day arrangement works at all. By that measure, a single read placed before the day starts leaves you exposed for one entire working day.
Placement dominates count. Take a firm whose correspondents are at their desks between 8:00 and 6:00. That’s ten hours of exposure to cover.
One read at 7:40 happens before any of it. The gap runs from 8:00 through to 7:40 the next morning, which is the full ten working hours.
Two reads at 7:40 and 8:45 sound like twice the coverage. They aren’t. The first gap is forty-five minutes. The second is the remaining nine and a quarter hours. Your exposure went from ten hours to nine and a quarter. You bought forty-five minutes, and you paid about 250 extra reads a year for it.
Two reads at 7:40 and 1:00 cut the same ten hours into five and five. Same count, same cost, half the exposure.
That’s the entire argument for spending ten minutes on this rather than four seconds. Which two hours you pick matters more than how many you pick. And the failure has a specific shape. People who add a second briefing tend to put it in the morning, because both slots feel like the part of the day when you deal with mail.
And the hours you’re covering belong to the people who write to you. If your consequential mail comes from two time zones east, their day opens and closes two hours before yours does. Move both reads earlier, rather than adding a third. It runs the other way too. A firm in Boston whose work comes out of California is exposed until 8:00 in the evening Eastern, and a morning read leaves that whole stretch open. Before you set anything, look at the send times on a month of the mail that actually mattered. It takes a few minutes, and it’s the one input to this decision that’s specific to you.
A federal rule that had to choose its hours
How many times a day to publish something that people plan around has been settled in writing by an American institution with rather more riding on it than your Tuesday.
On February 15, 2024, the Office of Management and Budget published a revised Statistical Policy Directive No. 3. It took effect four days later. It governs how federal agencies compile and release the series designated as principal Federal economic indicators. It opens by saying why a schedule exists at all. Because these numbers “have significant commercial value, may affect the movement of commodity and financial markets, or may be taken as a measure of the impact of government policies, public release must be prompt and according to an established, publicly available schedule” (checked September 6, 2026).
Then it makes three choices worth borrowing.
One or two, and no more. The release-schedule section reads: “Agencies shall establish and maintain one or two times of day for the release of their principal economic indicators and shall only release indicators at such designated times” (checked September 6, 2026). An agency producing many series, ready at many different hours, holds all of them for one or two times of day, rather than releasing each one as it becomes available. The directive leaves its reasoning unstated, and the shape of it is plain anyway. The value of a scheduled release is that everybody knows when it is. Every additional hour on the schedule dilutes exactly that.
The schedule is published, and published early. “The last report of each calendar year must contain the time and date of all reports in the upcoming year,” and “each release will include an announcement of the time and date of the next release” (checked September 6, 2026). A year ahead, and then again in every issue. Predictability is treated as a property of the thing rather than a courtesy to the reader.
It stays where it is. “Changes in the release schedule may occur only if special, unforeseen circumstances arise,” and any change must be announced and fully explained as soon as it is known to be unavoidable (checked September 6, 2026). This is the one people underestimate. A briefing hour you slide to 9:00 because Tuesday got away from you is a habit rather than a schedule, and the gap it leaves is a different gap from the one your exception list was written against.
Directive No. 3 governs federal statistical agencies and the economic series OMB designates. It exists to keep those releases policy-neutral, and to stop early access moving markets before everyone has the numbers. It’s silent on anybody’s mail. What transfers is the answer an institution has reached after thinking about this question since 1985: one or two fixed times, published a year out, and hard to move. If four briefings a day sounds thorough to you, that’s the counterweight.
One more line from it, because it names the trade the next section is about. “Deciding when to release a principal economic indicator requires the balancing of accuracy and timeliness” (checked September 6, 2026). Later is more complete. Later is also, past some point, too late to be worth having.
Working back from the first thing that cannot move
The clearest worked example of an hour chosen for the reader rather than for the writer is the monthly jobs report. Each Employment Situation release carries the line “Transmission of material in this news release is embargoed until 8:30 a.m. (ET)”, and the New York Stock Exchange gives its “Core Trading Session: 9:30 a.m. to 4:00 p.m. ET” (both checked September 6, 2026). Sixty minutes between them.
At 8:30 the numbers have already been ready for days, sitting under the embargo. The hour was chosen because the audience needs an hour with them before the moment it has to act. An hour is what the release gives.
That’s the calculation to run on your own morning, and it has three terms.
Find the first fixed point. The first thing on today’s calendar that moves only if you tell somebody. If your calendar holds nothing that qualifies, use the first hour you’re committed to another person. This is the deadline the briefing serves, and it’s a different hour from when you start work.
Subtract the read. The first chapter puts a good briefing at five minutes, and that’s the smallest of the three terms by a distance.
Subtract the longest thing it could ask you to do. Take the longest rather than the average, because the mornings that justify the whole arrangement are the ones where something has changed. Usually that means pulling a revised set of numbers, reading a document that landed after eleven at night, or getting hold of somebody.
The last of those is where the arithmetic goes wrong. An action that needs another person runs on their time. A phone call that takes six minutes needs ninety minutes of window, because the first two attempts find them on another call. If reaching a human being is among the plausible responses to your briefing, subtract ninety minutes rather than six.
So a ten o’clock with a forty-five-minute worst case means the briefing has to land by 9:10. If reaching somebody is among the plausible responses, by 8:30.
Then there’s a second edge, and it’s the one most people set last. A briefing composed at 5:30 describes a night. By 9:10 it can be missing three and a half hours, and those three and a half hours are precisely the ones holding your correspondents’ first mail of the day. So you read it at 5:30, and at 9:20 you open the inbox anyway. Now you’re running a briefing and a checking habit side by side, and paying for both.
The earliest useful hour is therefore about half an hour after the people who write to you start writing, whatever time you get up. For most firms that puts the whole window between about 8:00 and the first fixed point, which makes it narrower than people expect and later than they assume.
If what you want is the briefing before the household is awake, that’s a real preference, and it stands. All of this is an argument for knowing the price. You’ve chosen a stale read, so a second one in the late morning becomes part of the arrangement rather than an extra. The morning is the part the first read leaves for it.
The cutoffs your day is actually exposed to
The last read of the day comes out of the same arithmetic, pointed the other way. Find the earliest hour at which something can go wrong past fixing, and put a read far enough in front of it to do something about it.
The outer boundary of the payments day is public. The Federal Reserve states the Fedwire Funds Service hours (checked September 6, 2026). The “business day begins at 9:00 p.m. eastern time (ET) on the preceding calendar day and ends at 7:00 p.m. ET, Monday through Friday, excluding designated holidays.” Inside that day, “the deadline for initiating transfers for the benefit of a third party (such as a bank’s customer) is 6:45 p.m. ET each business day.”
That’s the system’s boundary, and yours sits earlier. Your own bank sets its own cutoff for accepting a wire, often earlier by hours, and that’s the one that binds you. Look it up rather than assuming, and write the number down. The rest of the list works the same way, and each answer is local to you. A client portal closes at end of day. Your payroll provider has a submission lead time. A courier has a last pickup. The local rules of one court set a filing window. What they have in common is that missing one costs a day, and a day is exactly the unit a once-daily briefing trades in.
Then the placement. Your last read has to sit ahead of the earliest of those cutoffs by at least the time it takes to act. If your bank stops accepting wires at 2:00 and assembling one takes twenty minutes, a read at 12:30 covers it. A read at 1:50 leaves you ten minutes for twenty minutes of work.
Which is why a second briefing is a midday one rather than an evening one. The worth of a read is the number of working hours downstream of it, and a 5:00 read sits at the end of them. It describes a day you’ve already finished changing, which is company more than information. An end-of-day read is a legitimate thing to want, and it’s a different object. It’s preparation for tomorrow, and it competes with your evening rather than your afternoon. What that costs is the subject of protecting your evenings, and it’s worth reading before you turn one on.
Two smaller placement notes. Put the midday read at the edge of a cleared block rather than inside one. A read at 11:00 in the middle of the only two uninterrupted hours you had is a self-inflicted version of the problem an empty afternoon is not a protected one describes. And put it after the arrival peak rather than before it, which for most firms means somewhere after 12:00. A briefing assembled at 11:15 out of a quiet morning is a read you’ll quietly stop doing.
While you’re placing it, check what a second read actually says, whatever software is producing it. The first chapter made the case for a count you can hold the thing to. Ninety-one arrived, six need you today. At 1:00 that sentence has two possible meanings, and only one of them is useful. A count against the whole day tells you where you stand overall. A count against the four hours since the last read tells you what changed. Read a couple of midday briefings and work out which one you’re being handed. A midday read that silently re-runs the morning is a read you’ll abandon inside a week.
Saturday, and the days that start differently
Two cases break a single fixed hour, and they break it in opposite directions.
The weekend is a Friday problem. Read Friday morning and again Monday morning and the gap runs to about seventy hours. Nearly all of it is harmless. The live part is Friday afternoon, which is the stretch of the week most likely to hold somebody clearing their desk before they leave and putting the result on yours. So what the weekend argues for is a late Friday read rather than a Saturday one, and it’s the single exception to everything above about evening reads. The usual objection to an evening read is the empty hours downstream of 5:00. On a Friday the next read is seventy hours away rather than fifteen.
If you want a weekend read as well, be clear about what it’s for. It’s for confirming that nothing is on fire, which is the specific reassurance that settles four furtive checks between Saturday lunch and Sunday night. One, in the morning, is cheap. Two a day across a weekend and you’re back at work.
Then the day that starts somewhere new each time. A fixed hour works when the first fixed point is roughly fixed. For anyone who travels, works on client sites, or has a Tuesday that starts at 7:00 and a Wednesday that starts at 11:00, it moves, and this is the case the first chapter left open.
Three ways out, each with a price.
Set the clock to your earliest start. Simple and predictable, and on a day that begins at 11:00 you’re opening something four hours old before you’ve done anything with it.
Anchor to the calendar instead of the clock, so the briefing lands a set interval before your first commitment. This is the answer that sounds best, and it quietly gives up the thing Directive No. 3 was protecting. The arrival hour becomes a surprise, to you as much as anyone, and on a day with an empty calendar it has no anchor at all.
Set the clock late and take two. The early read is thin and covers the early starts. The later one catches everybody else’s morning. For irregular days this is usually the right answer, and the interesting part is that the argument for it rests on the variance alone, rather than on how much mail you get.
A rule of thumb. If your start time moves by more than about two hours across a normal week, take two reads rather than hunting for the one correct hour. Place them near the ends of the range rather than in the middle of it.
There’s a fourth shape, worth naming because for some people it’s right. No schedule at all, and a briefing you ask for when you sit down. It fits a genuinely unpredictable week. Its cost is exact, though, and it’s real. The gap is now whatever length your day makes it, so the entire arrangement rests on the exception list, and that list has to be longer and better written than anyone else’s.
Testing the number against the list you brought
Now put the two together, which is what this chapter owes the last one.
Take the exception list and write a single number against each line. The number is how long the item could sit before sitting starts to cost you something.
Then sort into four piles.
Under two hours. These stay interrupts whatever you do. Covering a two-hour tolerance takes five reads a day, and at five reads a day you’ve rebuilt the inbox with better prose. Leave them on the list.
Two to five hours. A second read clears these, as long as every gap comes in under the tolerance. This is where the placement arithmetic from the first section stops being theoretical. A pair of reads at 7:40 and 8:45 leaves all of these standing. A pair at 7:40 and 1:00 clears most of them.
Same day, ahead of a cutoff. A read sitting before the cutoff minus the time to act clears these. For most firms this is the largest pile, and it usually decides whether a second read pays for itself.
Tomorrow would be fine. These belong off the list. If they’re still on it, the two-week audit from the last chapter is still waiting, and running it is cheaper than anything in this one.
Then the exchange rate, which is countable rather than felt. A second daily read costs about 250 reads a year at five minutes apiece, and the five minutes is the smaller half of it. A line on the exception list costs whatever it fires. Twice a month is 24 interruptions a year. Twice a week is about 100.
The two are priced differently, and the asymmetry is the point. A scheduled read is expected, it’s short, and it happens at an hour you chose. An interruption arrives unannounced and runs as long as it runs, which is why the last chapter put a false one at twenty minutes. So the exchange runs in your favor, up to a point. A second read that retires two or three lines firing weekly is clearly worth taking. A second read that retires one line firing monthly costs more than the line, so keep the line.
There’s a prior question, and it usually settles the whole thing. Are you already checking between briefings? If you are, you’ve already added a second read, and it works without the advantages a scheduled one has. It lands at a random hour, and it runs as long as it runs. It retires nothing from the list, because it surprised you too. Turning that into a setting names a read you’re already taking, and your day stays the same length. Why the pull is there, and what it’s actually responding to, is the checking habit itself.
And the ceiling. Somewhere around three or four scheduled reads spaced two hours apart, the arrangement has spent itself. The return was that the deciding happened once, before the day started. At four reads it happens four times, and the third of them falls at the hour your judgment is worst. If your arithmetic keeps coming out at four or more, the honest conclusion is that your desk is reactive. A reactive desk wants ranking applied to each message as it lands, rather than a schedule laid over the top. That’s the moment it lands. The broader version of the question, how many times a day to open an inbox at all, is when to check email, and if the answer keeps coming out high then that’s the better page to be reading than this one.
What Point has already done between reads
In Point both fields are settings rather than one fixed hour. The timing and the frequency are yours to place, so the schedule sits where your day needs it, which is what makes the arithmetic above something to act on rather than something to wish for.
A second read is cheap, and the reason is worth stating, because it’s what makes cadence a free choice rather than a costly one. Point summarizes every thread as it arrives, which is what lets you move through a full inbox at a glance in the first place, so a briefing is assembled out of records that already exist. A read at 1:00 is a second look at a ledger Point has kept continuously since the first one. The honest cost of adding it is your five minutes.
The two settings this series has now covered sit beside each other in one place. Where the interrupt threshold goes was the last chapter’s subject, and it belongs next to this one because of the trade between them. Raise the cadence and you can shorten the list. You see both at once, and you tune each one against the other.
The read also happens where you already are, since Point is the mail client rather than something that mails you a page. That matters more for a midday briefing than a morning one. The cost of a second read is mostly re-entry. A read that opens into the same feed you were working in costs less of that than one that means crossing to another window and finding your way back.
Two things sit outside the schedule. The first is authority. How far Point goes when Point acts is set separately for each kind of work, with three stops, and every kind starts at the middle one, so anything beyond that is a position you moved. Changing your briefing to twice a day changes when you’re told things, and what may be done on your behalf stays where you set it. The second is your mailbox. Whichever hours you pick, the mailbox underneath is still your existing Gmail or Microsoft 365 account, unmoved, on the address you already hand out. Everything Point does is the full inventory.
The run of ordinary days this assumed
Look at what you’ve actually built. An hour, a count, and a list, all three priced in rates: working hours of exposure, arrivals per hour, lines fired per year. Rates describe a system in a steady state, and every number in these three chapters has quietly assumed one, which is that yesterday looked like today and tomorrow will too.
That assumption holds for most of a year, and it’s the right thing to build on. It fails in one specific way, and in a way you’d be unlikely to guess. A busy day is only a higher rate, and the arithmetic survives it fine. What breaks it is a gap long enough that the things inside it start to act on each other. Come back from nine days away, and the message from day one has already been answered by the message from day four. The meeting you were preparing for has happened. The document you were chasing either arrived on day six or is now three weeks late. That morning wants something other than a longer briefing. Ranking it by consequence misses too, because half of it has been overtaken, and a report that solemnly lists overtaken things is worse than silence.
That’s a different object with a different job, and it’s where this series ends. Starting the day already sorted takes the morning that follows nine days at once, and the first morning of all, which turns out to have the same shape: nothing before it to be measured against, and the whole mailbox arriving in one piece.