Not every task arrives in a message. Some of them you think of yourself, halfway through reading something else, and there is nothing to lift them out of because they were never written down by anybody. Those have to be typed. What the typing costs is the whole question, because a capture expensive enough gets skipped, and the one you skip is the one you lose.
- The tasks nobody sent you are the ones no amount of reading your mail will find. They are not in the mail.
- The cost that decides everything is the gap between thinking of the thing and being able to start typing it. Applications and forms are what fill that gap.
- Measured, people capture in seconds and in fragments, and they pick the tool by how fast it opens rather than by what it can do.
- Speed decides whether the note exists at all. Asked where a thought would have gone without something quick to catch it, people mostly answer: nowhere.
The tasks that were never in a message
An ask sitting inside somebody’s paragraph can be found and dated by software reading your mailbox, and email task management is about exactly that. This page is about the remainder, which is larger than it looks.
Some of it is obvious. A promise you made on a call. Something agreed in a meeting that never generated a single email. A supplier who told you in person that prices go up in October.
The interesting kind is quieter. A message arrives and prompts a task it does not contain. You open a client’s mail about the March job and remember, while reading it, that nobody ever raised the invoice for February. Nothing in that message mentions February. A reader looking for obligations in the sender’s words would be right to find none, because the sender put none there. The task came from you, cued by the thread, and the thread is where you are standing when it arrives.
That timing is not bad luck. Those thoughts surface because your attention is on the subject, which means they show up while you are busy with something related, and the busy part is what makes them expensive to record.
There is a decent measure of how much of this goes on. In Information Scraps, published in ACM Transactions on Information Systems in 2008, Michael Bernstein, Max Van Kleek and David Karger at MIT with mc schraefel at Southampton interviewed 27 knowledge workers at five organizations, sitting with each at their own desk and asking about information they had that was not formally recorded in a proper place. They indexed 533 pieces of it, physical and digital. The single most common type was the to-do, at 92 instances, ahead of meeting notes, contact details and how-to guides.
Those to-dos were on paper notebooks and Post-its, in text files, inside emails people had sent to themselves, and in one photographed case written on a participant’s hand. Every one of those 27 people had proper software available. The to-dos were somewhere else.
What a capture actually costs
The paper is precise about which interval matters, and it is not the typing. Navigational and cognitive costs from launching or switching applications, the authors write, “contribute significantly to the time elapsed between the moment an individual forms an intention of writing something down and the moment they can actually start doing so”.
That is the number a working person is really pricing. One of their participants, a finance worker, put it in one line: “If it takes three clicks to get it down, it’s easier to e-mail.”
Forms cost more than clicks. A manager in the same study kept notes on Post-its stuck to his mobile phone so he could type them up later, rather than entering them on the phone, which was perfectly capable of taking them. Asked why, he said that starting in the desktop application “forces me to make a type assignment, assign a category, set a deadline, and more”. The authors generalise it: entering something into a structured form or application “could often double or triple the time it would take to simply type the information”. Their design conclusion names the culprits as upfront decisions, and lists three of them, categorising the item, choosing a reminder time, and setting parameters “ultimately unimportant to the captured information”.
The friction was real enough to shape their own method. They ruled out a diary study, they explain, because the effort of writing in a diary would have suppressed the very behaviour they were trying to observe: participants “may have simply chosen not to record the scrap to avoid the effort”.
So people substitute, and every substitute is cheaper at the moment of capture and dearer afterwards. Paper, which is instant and gets lost. A text file, which is instant and is not a list you work from. Your own memory, which is free for about ten minutes. And the inbox, which is where the largest share of it goes: the same paper, summarising earlier research, notes that nearly a third of all archived email turns out to have been sent by the owner to themselves.
Every one of those is a person paying later so they can keep working now. That is not a discipline problem. It is an accurate reading of what the moment costs.
Why five seconds is a threshold
The same group built the smallest capture tool they could and watched what happened. Note to Self, presented at the 2009 conference on human factors in computing systems by Max Van Kleek, Michael Bernstein, Katrina Panovich, Greg Vargas, David Karger and mc schraefel, gave 42 people a plain text box in their browser sidebar for ten days. It was deliberately restricted to what every note tool has, making a note, deleting one and searching by keyword, plus a pop-up input box you could call up by hotkey without opening the sidebar at all, so that catching a thought did not require looking away from whatever had produced it.
Three numbers describe how it got used. 30% of notes were captured in five seconds or less. 50% in ten seconds or less. 95% within two minutes. The authors’ reading of that distribution is the sentence worth keeping: “The speed with which notes were captured indicates that every second counts.”
Asked why they had put a particular note there rather than somewhere better suited, people named quick capture more often than anything else, on 35% of the 290 notes polled, ahead of browser integration at 18%, being able to see the notes at 13% and search at 7%. And 69% of what they filed, they classified as to-dos. Given a box that did nothing except open fast, people used it as a task list.
Then the finding that settles the argument. Asked where a note would have ended up without the tool, several answered that it would not have existed. “I probably would not have taken a note at all, and I probably would have forgotten to do it,” said one. Another had been in the habit of using paper “that would then get lost”.
Forty-two self-selected volunteers over ten days, with a small payment for the prompted notes, is not a seven-person accounting firm, so read the shape rather than the decimals. The shape is the useful part, and it is this: capture speed does not make an existing habit more pleasant. It decides whether the habit exists.
Which reframes what five seconds is a claim about. It is not a boast about typing. It is the difference between a system that holds on the Tuesday when three things go wrong and one that quietly stops being fed that morning and is never trusted again. The failure that kills one of these is rarely a review you skipped. It is an intake that stopped.
What a five-second note actually says
Capture that fast is capture that compresses. In the same ten days, the median note ran 29 characters and seven words. 7% were a single word, 43% were five words or fewer, and 78% carried no punctuation at all. Asked afterwards what one of those single words meant, a participant explained that “jhsieh” stood for needing to contact a person as soon as there was time.
The obvious worry follows immediately, and somebody checked it. Of the 533 scraps indexed in the interview study, exactly one was left with its owner unable to say what it meant. People are good at reading their own shorthand. But the authors are careful about the size of that claim, and so should anyone quoting it be: they tested memory for the gist, not for the particulars, and note that specific details “would likely have fared more poorly”.
That is the honest split. What survives compression is what the note is about. What does not survive is the part you knew at the time and did not bother to type, which is usually the when, sometimes the which, and occasionally the whether-you-already-did-it.
Two more numbers say the same thing from the other end. Three-quarters of notes were never edited after they were captured, and another 19% were edited exactly once. Whatever went in during those five seconds is, in practice, the finished article. So the fix is not to make people write more, which is just the form again wearing a different hat. The fix is for the five seconds to buy a task rather than a fragment: the sentence you typed carries the date inside it, and something reads it out of the sentence instead of asking you for it in a field.
Where the task came out of somebody else’s message the missing context is their words, and keeping the link back to them is the repair; that case belongs to email task management. Here you are the author, so the risk is the reverse one, and the repair is a due date you did not have to stop and choose.
There is a second way this goes wrong, and it has nothing to do with wording. A fast box that is its own private scratchpad has not given you a task list. It has given you a second one to reconcile, which is the failure one surface, not six apps sets out at length. Speed only counts if the note lands where you already work.
What speed does not buy
It does not decide anything. Five seconds buys you permission to stop carrying the thought around. What to do about it is untouched, and pretending otherwise is how people end up with a beautifully captured backlog.
It does not make the list shorter. A cheaper intake means more gets in, which is the direction you want and is not free. A list that finally contains everything is a list that has to be worked through, and sorting what matters from what merely has a date on it is a separate skill: important versus urgent is about that split.
It cannot catch what you never thought of. This path runs on your noticing. It is the complement to the automatic one rather than a replacement for it, and between them they still miss the thing nobody said and nobody remembered.
It does not fix a bad moment. Some thoughts arrive with your hands on a steering wheel or a client mid-sentence. The best a capture tool can do there is be reachable without looking, and the worst it can do is require a screen.
How Point takes a task
In Point, the manual path starts where you already are. You write the task as one ordinary sentence, in the same place you were reading or working, and it becomes a real task with a due date and your name on it. It is confirmed there and then, with a link straight to it, so the thread in front of you is still the thread in front of you. Opening the task is a choice rather than a consequence.
Where it lands is the part that matters more than the speed. It joins one list alongside the tasks lifted out of your mail and the tasks belonging to any other business you run, rather than starting a private pile of its own. One click marks it done, and what you have finished is counted rather than only what is still waiting, which is a small thing that changes how a list feels by Friday.
When your hands are full you can hold to talk and get the same result you would get by typing. And because a task typed in five seconds is a task you will later have to find, one search runs across your mail, tasks, calendar and files together, so finding it does not begin with remembering where you put it.
The wider controls sit around all of this. Each kind of action has its own setting for how far Point may go alone, from suggesting through preparing and waiting on you through to handling it outright, with the middle one in place until you move it. Everything done for you is written down in order, and an entry in that record is where reversing it starts from. The rest of what Point does with your mail is set out on the benefits page.
Common questions
How is this different from a task Point finds in an email?
Only in who noticed. Both end up in the same list with the same dates, and the automatic path is the one doing most of the volume, because most of what a small firm owes did arrive in writing. The manual path exists for the case no reader can cover: a task that no message contains, because the person who thought of it was you, and until you typed it the only copy was in your head.
Will a task I typed in five seconds still make sense next week?
Mostly yes, and the research is reassuring on that point, because you are reading your own shorthand rather than somebody else’s. What decays is the detail you did not type because it was obvious at the time. So spend the extra second on the noun rather than the sentence, which client and which invoice, and let the date come from the words you were going to write anyway.
Is it not quicker to just remember it?
For the next ten minutes it genuinely is, and that is exactly the trade people make all day. The bill arrives later, at the moment you needed the thought and it was not there. Nothing argues a busy person out of that trade. The only thing that shifts it is making the writing cost less than the remembering feels like it costs, which is the entire design problem.
Does a faster way to add tasks just leave me with a longer list?
At first, yes, and that is the correct direction. The list grows because it is becoming accurate, not because more work appeared. What used to happen instead was that the surplus lived in your head and in a drawer, where it looked like less and was not. A longer honest list moves the difficulty from remembering to deciding, which is the better place for it.
What about a task I think of when I am not at a screen?
Hold to talk and say it, which is the same capture without the keyboard. Where there is no device at all, on a site visit or walking back to the car, nothing catches it but you, and the realistic aim is to shorten how long you have to carry it rather than to pretend the gap is closed. A promise made in a corridor is still the hardest thing in this whole subject.
The short version
The path where a message carries the task is the one worth automating first, and it is not the whole of what you owe. The rest you think of yourself, usually while doing something else, and it survives only if writing it down is cheaper than not bothering. That is a measurable thing rather than a matter of character: people capture in seconds, in fragments, and they choose the fastest surface over the best-designed one every time. Take away the second application and the form and the upfront decisions, and what is left is a sentence, a date read out of it, and one list you actually work from.
Which thoughts turn up on that list depends entirely on the trade. They read one way in an accounting practice and another in a consultancy, and there are eight versions written out.