Mail arrived while you were asleep, and each message was ranked as it landed rather than when you sat down. So the first morning is short: read the top of the feed, spend a minute checking behind it in the one place a mistake could hide, and then work out what to do with what is left of the hour.
Yesterday you connected a mailbox and then deliberately left everything alone. Nothing has been sent since and nothing has been filed without you, because every kind of work starts at review and you moved none of them. The only thing that happened overnight was a judgment about order, which makes this morning a narrow test with one question in it: did anything that needed you today end up somewhere you would not have looked?
That question has a one minute answer. Most of this page is about what to do once you have it.
What is actually on the screen
Open Point and the night is already in three groups, with the header keeping the score: what needs you now, what can wait, and what is just chatter. Each thread is one short line written by Point, with a dot and a time beside it. The dot means this one needs you today.
The score is the first surprise, and it usually goes the same way. A night’s worth of mail arrived and a handful of it is being called today’s work. Nobody’s inbox has ever told them that before, and the reasonable response to being told it by software you met yesterday is doubt.
Two smaller things about the screen, so they do not distract you from that.
The order inside the top group will not be the order you would have chosen, and that is not the fault worth hunting for. Every message in there is one you were going to handle today. Whether the sign-off sits above the payment query is a preference, and the reason those two are hard to rank against each other at all is that importance and urgency are separate measurements pulling in different directions, taken apart properly in why important and urgent are not the same.
And if you kept the morning briefing, a short account of the day will have arrived at whatever hour you set it for, counting what needs you against what is on the calendar. Read it, then judge the feed rather than the briefing. The briefing is a report of a judgment; the feed is the judgment, and the judgment is what is on trial today. Everything Point does covers the rest of what turns up with the morning, for the week when you are curious rather than suspicious.
Check behind it in the one place that matters
Yesterday ended with a prediction about what you would do next: read the sorted feed, then open Gmail anyway and scroll the real list to the bottom to see what had been kept from you. Do the checking. Do not do it that way.
Scrolling the whole mailbox is the old morning wearing a new hat. You are reading a hundred rows in arrival order looking for nothing in particular, which is the exact activity that let things slip in the first place, and it produces nothing you can use at the end of it. You reach the bottom, nothing jumped out, and all you have learned is that nothing jumped out.
There is a version that takes a minute and is complete.
Connecting Gmail put three labels on your real mailbox, and every triaged message carries the one it earned. You have already read Now, because that was the top of the feed. Later is a deferral rather than a disappearance, sitting in your mailbox all day, costing you nothing to open at eleven o’clock. Which leaves one group. Other is the only place where something was decided against you seeing this morning, so it is the only place a mistake can be, and it is a label you can click.
Click it, and read the senders rather than the messages. You are looking for one thing: a name you would have opened. On most mornings that takes well under a minute, because what is in there is newsletters, receipts, notifications and threads you were copied into, and a name that does not belong in that company is visible at a glance rather than by reading.
That is a complete audit rather than a sample, which is what scrolling never was. Every message Point decided you did not need is in that one view, and when you close it you have looked at all of them.
Do it in Gmail rather than in Point this morning. Outlook is the same check, with categories doing the labels’ job. Point has its own version, one tap that dissolves the groups and lays the feed out in plain time order, newest first, and it is genuinely useful later on. But a check on the sorting is worth more when it is run from the app that did not do the sorting, and today you have less reason to take anything on trust than you will ever have again.
One thing that looks like a miss and is not. If a message is in none of the three groups and not in your inbox either, an old rule of your own archived it on arrival before anything else saw it, which is the trap yesterday’s page told you to expect. Your filter is the first suspect and the software is the second.
What counts as a miss and what is just noise
Three things get called failures on a first morning, and only one of them is.
An order you disagree with inside the top group is not a miss. It cost you the time it took to read four lines in a sequence you would not have picked.
Something in the top group that did not deserve to be there is a miss, and a cheap one. A newsletter that reads like a client, a payment notification that looks like a question. Note it and leave it alone. A false alarm at the top costs you a glance. The other direction is the one that costs you a client.
Something sitting in Other that you would have answered today is the real thing, and the number worth carrying out of this morning is how many of those there were. It should be zero, and on most first mornings it is.
If it is not zero, the message it happened to is usually one of two kinds, and both are worth knowing in advance rather than discovering.
The first is somebody writing to you for the first time. A ranking that leans on your history with a sender has no history to lean on the first time somebody new writes, so a first approach is at once the message you can least afford to lose and the one there is least to go on with. What a model can genuinely judge about your inbox is where that limit comes from and what the accuracy underneath it is worth.
The second is the short one from somebody you know that reads like nothing at all. “Can you give me a ring today.” No real subject, no question in it, no date, and everything at stake sitting in the part that was never written down. You would have known what it meant, because you know what happened last week. Nothing that read only the message could have.
Neither of those is a reason to stop, and neither is a fault you can settle this morning. They are the two places to point your eyes tomorrow.
The fix is a sentence rather than a settings screen
Say you found one. Here is what to do with it today, and more usefully, what not to.
If the problem is a person, star them. One star on a contact, your judgment entered directly rather than inferred from anything, and everything that person sends scores higher from then on. It is the shortest path there is between noticing and fixing. Do it for the one person you just found and then stop, because the proper pass over your contacts is worth having a week of evidence behind it, and it comes later in this fortnight for that reason.
If the problem is that Point’s account of a thread was wrong rather than its position, say so in a sentence, in your own words, on the thread. That correction is kept as a standing rule rather than as an edit to one summary, and the next thing that sender writes is summarized the new way. Nobody expects you to phrase it carefully. What a summary has to carry, and why yours is allowed to differ from anybody else’s, is a plain summary of every email.
What not to do is move the dial. The temptation runs in both directions this morning: a miss makes you want to pull something back, and a feed that did something impressive at seven o’clock makes giving it more rope feel like the natural reward. Both are the same error. The dial sets how much of the work Point does on its own, and the ranking is not one of the settings on it. This morning taught you nothing about filing, drafting or scheduling, and moving one of those in response is answering a question nobody asked, at the cost of being able to explain anything that happens on Friday.
While you are resisting that, find the record once. It is the list of what was done on your behalf, in order and in plain words, and today it is close to empty, which is the correct and slightly flat result of having left everything at review. Knowing where it lives while it still says nothing is better than going looking for it in a fortnight when you badly want it to say something. The record of what was done, and the way back is what collects in there and what the entries are for.
The hole where the sorting used to be
It is twenty past nine and you have finished.
This is the strangest thing about a first morning with a sorted inbox, and nobody warns you about it. The work you did was real: four threads, three answered, one set aside. What is missing is everything that used to be around it. The half hour of ranking and skimming and deciding was never only ranking and skimming and deciding. It was how you arrived at work. It was the thing you did with a coffee while your head came online, and it reliably produced the feeling of having started, which turns out to be a different thing from having started.
Take it away and the first real task of the day is standing there at twenty past nine with no warm up in front of it. Most people meet that as unease rather than as a gain, and then they cure the unease in one of two ways, both of which cost you the fortnight.
The first is going back to the inbox and reading it anyway. Not checking, which you have already done properly and which took a minute. Reading. Working down the list, opening things, feeling busy, quietly re-earning the sensation the morning used to hand over for free. Do that daily for two weeks and you will finish genuinely unable to say whether the software helped, because you spent the time it gave back on the activity it took away.
The second is tidying. Rebuilding rules, rearranging folders, walking through settings, deciding that today is the day to do something about the archive. Same shape, same cause, and it feels more legitimate because it looks like setup. There is a real pruning exercise waiting at the end of this fortnight, and it will be a much better informed one then.
What to do instead is decide, the night before, what the first real piece of work is. Write it on something. The time is coming back whether or not you have somewhere to put it, and an hour that arrives without a destination gets spent on the inbox by default.
Two numbers worth writing down
Before this morning is out of your head, write down two things somewhere you will find them in a fortnight.
The first is the clock time you stopped dealing with email and started on something else. Not how long it felt. The time on the clock.
The second is the count out of the Other label: how many messages were in there that you would have answered today. Most likely zero, possibly one.
Neither number means anything this morning, and that is exactly why they are worth writing down. The first is the baseline the whole fortnight gets measured against. The second is the only safety question that matters, asked on a day you are still checking rather than in three months when you have quietly stopped. In a fortnight you will have the same two numbers again and a straight comparison, which is worth considerably more than a memory of how the first week felt.
There is a third number from this morning that you cannot write down, because you were not counting while it was happening.
You opened almost every thread. You read the short line Point wrote, then opened the mail underneath to see whether the line was right, then agreed with the line and closed it again. Twenty times, thirty, on messages whose shape you had already been told. That was the correct behaviour for day two. It is also the whole of the time you did not save this morning.
Tomorrow the inbox will be sorted before you get there, and the check will take you forty seconds instead of a minute, because you will know where to look. The question stops being whether the order is right. It becomes whether the line is enough to act on without opening the thing underneath it, and giving that up is harder than giving up the sorting was.