I’ve been using the Internet for a good 25 years now, and I’ve been lucky
enough to have some perspective dating back farther than
that. The common refrain for my entire tenure here:
We all get too much email.
A New, New, New, New Hope
Luckily, something is always on the cusp of replacing email. AOL instant
messenger will totally replace it. Then it was blogging. RSS. MySpace. Then
it was FriendFeed. Then Twitter. Then Facebook.
Today, it’s in vogue to talk about how Slack is going to replace email. As
someone who has seen this play out a dozen times now, let me give you a little
Slack is not going to replace email.
But Slack isn’t the
here, either. It’s just another communication tool.
The problem of email overload is both
If the problem were really with “email”, then, presumably, one of the nine
million email apps that dot the app-stores like mushrooms sprouting from a
globe-spanning mycelium would have just solved it by now, and we could all
move on with our lives. Instead, it is permanently in vogue to talk about
how overloaded we all are.
If not email, then what?
If you have twenty-four thousand unread emails
in your Inbox,
what you’re bad at is not email, it’s transactional interactions.
Different communication media have different characteristics, but the defining
characteristic of email is that it is the primary mode of communication that we
use, both professionally and personally, when we are asking someone else to
perform a task.
Of course you might use any form of communication to communicate tasks to
another person. But other forms - especially the currently popular real-time
methods - appear as a bi-directional communication, and are largely immutable.
Email’s distinguishing characteristic is that it is discrete; each message is
its own entity with its own ID. Emails may also be annotated, whether with
flags, replied-to markers, labels, placement in folders, archiving, or
deleting. Contrast this with a group chat in IRC, iMessage, or Slack, where
the log is mostly unchangeable, and the only available annotation is “did
your scrollbar ever move down past this point”; each individual message has
only one bit of associated information. Unless you have catlike reflexes and
an unbelievably obsessive-compulsive personality, it is highly unlikely that
you will carefully set the “read” flag on each and every message in an extended
All this makes email much more suitable for communicating a task, because the
recipient can file it according to their system for tracking tasks, come back
to it later, and generally treat the message itself as an artifact. By
contrast if I were to just walk up to you on the street and say “hey can you do
this for me”,
you will almost certainly just forget.
The word “task” might seem heavy-weight for some of the things that email is
used for, but tasks come in all sizes. One task might be “click this link to
confirm your sign-up on this website”. Another might be “choose a time to get
together for coffee”. Or “please pass along my resume to your hiring
department”. Yet another might be “send me the final draft of the Henderson
Email is also used for conveying information: here are the minutes from that
meeting we were just in. Here is transcription of the whiteboard from that
design session. Here are some photos from our family vacation. But even in
these cases, a task is implied: read these minutes and see if they’re accurate;
inspect this diagram and use it to inform your design; look at these photos and
just enjoy them.
So here’s the thing that you’re bad at, which is why none of the fifty
different email apps you’ve bought for your phone have fixed the problem: when
you get these messages, you aren’t making a conscious decision about:
- how important the message is to you
- whether you want to act on them at all
- when you want to act on them
- what exact action you want to take
- what the consequences of taking or not taking that action will be
This means that when someone asks you to do a thing, you probably aren’t going
to do it. You’re going to pretend to commit to it, and then you’re going to
flake out when push comes to shove. You’re going to keep context-switching
until all the deadlines have passed.
In other words:
The thing you are bad at is saying ‘no’ to people.
Sometimes it’s not obvious that what you’re doing is saying ‘no’. For many of
us — and I certainly fall into this category — a lot of the messages we get are
vaguely informational. They’re from random project mailing lists, perhaps
they’re discussions between other people, and it’s unclear what we should do
about them (or if we should do anything at all). We hang on to them (piling up
in our Inboxes) because they might be relevant in the future. I am not
advocating that you have to reply to every dumb mailing list email with a
5-part action plan and a Scrum meeting invite: that would be a disaster. You
don’t have time for that. You really shouldn’t have time for that.
The trick about getting to Inbox Zero is not in somehow becoming an
email-reading machine, but in realizing that most email is worthless, and
that’s OK. If you’re not going to do anything with it, just archive it and
forget about it. If you’re subscribed to a mailing list where only 1 out of
1000 messages actually represents something you should do about it, archive all
the rest after only answering the question “is this the one I should do
something about?”. You can answer that question after just glancing at the
subject; there are times when checking my email I will be hitting “archive”
with a 1-second frequency. If you are on a list where zero messages are ever
interesting enough to read in their entirety or do anything about, then of
course you should unsubscribe.
Once you’ve dug yourself into a hole with thousands of “I don’t know what I
should do with this” messages, it’s time to declare email
bankruptcy. If you have
24,000 messages in your Inbox, let me be real with you: you are never, ever
going to answer all those messages. You do not need a
to tell you exactly how many messages you are never going to reply to.
We’re In This Together, Me Especially
A lot of guidance about what to do with your email addresses email overload as
a personal problem. Over the years of developing my tips and tricks for
dealing with it, I certainly saw it that way. But lately, I’m starting to see
that it has pernicious social effects.
If you have 24,000 messages in your Inbox, that means you aren’t keeping track
or setting priorities on which tasks you want to complete. But just because
you’re not setting those priorities, that doesn’t mean nobody is. It means
you are letting
availability heuristic -
whatever is “latest and loudest” - govern access to your attention, and
therefore your time. By doing this, you are rewarding people (or #brands) who
contact you repeatedly, over inappropriate channels, and generally try to flood
your attention with their priorities instead of your own. This, in turn,
creates a culture where it is considered reasonable and appropriate to assume
that you need to do that in order to get someone’s attention.
Since we live in the era of subtext and implication, I should explicitly say
that I’m not describing any specific work environment or community. I used to
have an email startup, and so I thought about this stuff very heavily for
almost a decade. I have seen email habits at dozens of companies, and I help
people in the open source community with their email on a regular basis. So
I’m not throwing shade: almost everybody is terrible at this.
And that is the one way that email, in the sense of the tools and programs we
use to process it, is at fault: technology has made it easier and easier to ask
people to do more and more things, without giving us better tools or training
to deal with the increasingly huge array of demands on our time. It’s easier
than ever to say “hey could you do this for me” and harder than ever to just
say “no, too busy”.
Mostly, though, I want you to know that this isn’t just about you any more.
It’s about someone much more important than you: me. I’m tired of sending
reply after reply to people asking to “just circle back” or asking if I’ve seen
their email. Yes, I’ve seen your email. I have a long backlog of tasks, and,
like anyone, I have trouble managing them and getting them all done, and I
frequently have to decide that certain things are just not important enough to
do. Sometimes it takes me a couple of weeks to get to a message. Sometimes I
never do. But, it’s impossible to be mad at somebody for “just checking in”
for the fourth time when this is probably the only possible way they ever
manage to get anyone else to do anything.
I don’t want to end on a downer here, though. And I don’t have a book to sell
you which will solve all your productivity problems. I know that if I lay out
some incredibly elaborate system all at once,
it’ll seem overwhelming. I know that if I point you at some
amazing gadget that helps you keep track
of what you want to do, you’ll either balk at the price or get lost fiddling
with all its knobs and buttons and not getting a lot of benefit out of it. So
if I’m describing a problem that you have here, here’s what I want you to do.
Step zero is setting aside some time. This will probably take you a few hours,
but trust me; they will be well-spent.
First, you need to declare email bankruptcy. Select every message in your
Inbox older than 2 weeks. Archive them all, right now. In the past, you might
have to worry about deleting those messages, but modern email systems pretty
much universally have more storage than you’ll ever need. So rest assured that
if you actually need to do anything with these messages, they’ll all be in
your archive. But anything in your Inbox right now older than a couple of
weeks is just never going to get dealt with, and it’s time to accept that fact.
Again, this part of the process is not about making a decision yet, it’s just
about accepting a reality.
One extra tweak I would suggest here is to get rid of all of your email
folders and filters. It seems like many folks with big email problems have
tried to address this by ever-finer-grained classification of messages, ever
more byzantine email rules. At least, it’s common for me, when looking over
someone’s shoulder to see 24,000 messages, it’s common to also see 50 folders.
Probably these aren’t helping you very much.
In older email systems, it was necessary to construct elaborate header-based
filtering systems so that you can later identify those messages in certain
specific ways, like “message X went to this mailing list”. However, this was
an incomplete hack, a workaround for a missing feature. Almost all modern email
clients (and if yours doesn’t do this, switch) allow you to locate messages
like this via search.
Your mail system ought to have 3 folders:
- Inbox, which you process to discover tasks,
- Drafts, which you use to save progress on replies, and
- Archive, the folder which you access only by searching for information you
need when performing a task.
Getting rid of unnecessary folders and queries and filter rules will remove
things that you can fiddle with.
Moving individual units of trash between different heaps of trash is not being
productive, and by removing all the different folders you can shuffle your
messages into before actually acting upon them you will make better use of your
time spent looking at your email client.
There’s one exception to this rule, which is filters that do nothing but cause
a message to skip your Inbox and go straight to the archive. The reason that
this type of filter is different is that there are certain sources or patterns
of message which are not actionable, but rather, a useful source of reference
material that is only available as a stream of emails. Messages like that
should, indeed, not show up in your Inbox. But, there’s no reason to file them
into a specific folder or set of folders; you can always find them with a
Make A Place For Tasks
Next, you need to get a task list. Your email is not a task list; tasks are
things that you decided you’re going to do, not things that other people have
asked you to do. Critically, you are going to need to parse e-mails into
tasks. To explain why, let’s have a little arithmetic aside.
Let’s say it only takes you 45 seconds to go from reading a message to deciding
what it really means you should do; so, it only takes 20 seconds to go from
looking at the message to remembering what you need to do about it. This
means that by the time you get to 180 un-processed messages that you need to do
something about in your Inbox, you’ll be spending an hour a day doing nothing
but remembering what those messages mean, before you do anything related to
actually living your life, even including checking for new messages.
What should you use for the task list? On some level, this doesn’t really
matter; task management is deeply personal, and if you don’t have one already,
you’ll probably need to experiment with a few options to see what works best
However, if you’re struggling with email overwhelm, and you don’t have a task
system you trust already now is not the time to see what works best. You
need something that is acceptable enough that you can use it to get started
Experimentation with different tools can itself be a form of procrastination,
so you need something right now, and probably something you can start using
without worrying about whether it’s worth paying for. Since you probably use
let me save you from analysis paralysis and tell you what you should use:
Google Tasks. In order to create tasks from emails, you just view the message
by hitting enter, then hit
T and type in a brief description of the
task, then hit the somewhat less obvious
⌘⌥, on the Mac)
on the Mac) to go back to the message list. You can open the task list at any
time with the shortcut
g k, so it’s always convenient. There’s a handy
so you can easily see your tasks on the go as well. While there are more
powerful tools available, Google Tasks is a great “good enough” option that
there’s really no reason not to try.
If you’re in the Microsoft ecosystem, Outlook has similar built-in
and while I haven’t tried it, I’m sure it’s an equally good place to start.
If you’re looking to make a bigger investment, or you’ve already tried these
entry-level options and they didn’t work for you, there are a plethora of
other options. Personally, I use—and heavily
recommend—OmniFocus. It has a handy
Service that you can easily bind to a keyboard
which works to turn tons of different things into tasks, including emails.
The mac has a real embarrassment of riches here, with OmniFocus’s main rival
being Things, which has similar
other platforms, I have heard very good things about
Todoist, which also has direct support for a number of
email clients. If you
enjoy customization, Remember the Milk also supports directly creating tasks
Oddly, although Apple’s Mac Mail.app has excellent support for sharing
references to messages with other applications, iOS’s MobileMail is severely
lacking in this department. If you need to do email processing on iOS, I’d
avoid the built-in mail app, and either use the native Gmail app, which has an
“add to tasks” menu option, or get Airmail which has
direct support for pretty much every task-management app under the
As you evaluate different software and services, it’s easy to get lost in
aesthetics or try out tons of features, but I think it’s important to keep
these basics in mind:
- You have to be able to add to it with minimal friction. “Minimal” here is
highly dependent upon your personal tolerance for fiddling with stuff. If
copying and pasting the contents of emails into your tasks doesn’t
particularly bother you, then even a text file could work. You may have
noticed that above I was really focusing on how each system can directly
produce a task from an email with no additional work, and that is because
for many people—myself included—even small bits of friction discourage us
ever so slightly from using the system, which makes it borderline useless.
Note how many tasks you’re creating and whether you’re actually completing
- You have to be able to access it wherever you do stuff, whether personal
stuff or work stuff. This means, among other things, you need something
that’s going to sync to all of your devices. Thankfully in the modern
cloud-connected world, almost every serious contender allows for this.
- You have to be able to get back to the emails from this system. In most
cases it’s enough for the task list to support hyperlinks of some kind, or
just search for the subject line in quotes, but again, this depends on your
threshold for fiddling. For me, I need a one-click way to find that
original message so I can immediately hit “reply”.
Again, I would recommend that you just jump in for starters to the most
convenient option; you can always export your data
later. I know that
some of you are going to need to do the fiddly tools-setup thing regardless, so
hopefully the criteria above can help you at least navigate that process more
Turn Messages Into Tasks
The next step - and this is really the first day of the rest of your life -
start at the oldest message in your Inbox, and work forward in time.
Look at only one message at a time. Decide whether this message is a
meaningful task that you should accomplish. Again, some practical advice if
you’re in the most common case of using Gmail, you can quickly access the
oldest messages by clicking on the thing that says “1-50 of 37340” on the
top-right of the message list and selecting “Oldest”.
If you decide a message represents a task, then make a new task on your
task list. Decide what the task actually is, and describe it in words; don’t
create tasks like “answer this message”. Why do you need to answer it? Do you
need to gather any information first?
If you need to access information from the message in order to accomplish the
task, use the mail-linking functionality of your tasks app that I described
above, or at least copy a link or note an easily-searchable feature of the
message, but in the worst case, following the guidelines above about
eliminating unnecessary folders and filing in your email client, just put a
hint into your task list about how to search for the message in question
Once you’ve done that:
Archive the message immediately.
The record that you need to do something about the message now lives in your
task list, not your email client. You’ve processed it, and so it should no
longer remain in your inbox.
If you decide a message doesn’t represent a task, then:
Archive the message immediately.
Do not move on to the next message until you have archived this message. Do
not look ahead. The presence of a message in your Inbox means you need to
make a decision about it. Follow the
touch-move rule with your
email. If you skip over messages habitually and decide you’ll “just get back
to it in a minute”, that minute will turn into 4 months and you’ll be right
back where you were before.
Circling back to the subject of this post; once again, this isn’t really
specific to email. You should follow roughly the same workflow when someone
asks you to do a task in a meeting, or in Slack, or on your Discourse board, or
wherever, if you think that the task is actually important enough to do. Note
the slack timestamp and a snippet of the message so you can search for it
again, if there is a relevant attachment. The thing that makes email different
is really just the presence of an email box.
Banish The Blue Dot
Almost all email clients have a way of tracking “unread” messages; they
cheerfully display counters of them. Ignore this information; it is useless.
Messages have two states: in your inbox (unprocessed) and in your archive
(processed). “Read” vs. “Unread” can be, at best, of minimal utility when
resuming an interrupted scanning session. But, you are always only ever
looking at the oldest message first, right? So none of the messages below it
could possibly have been read yet anyway...
As you try to start translating your flood of inbound communications into an
actionable set of tasks you can actually accomplish, you are going to notice
that your task list is going to grow and grow just as your Inbox was before.
This is the hardest step:
Decide you are not going to do those tasks, and simply delete them. Sometimes,
a task’s entire life-cycle is to be created from an email, exist for ten
minutes, and then have you come back to look at it and then delete it. This
might feel pointless, but in going through that process, you are learning
something extremely valuable: you are learning what sorts of things are not
actually important enough for you to do.
If every single message you get from some automated system provokes this kind
of reaction, that will give you a clue that said system is wasting your time,
and just making you feel anxious about work you’re never really going to get
to, which can then lead to you un-subscribing or filtering messages from that
Tasks Before Messages
To thine own self, not thy Inbox, be true.
Try to start your day by looking at the things you’ve consciously decided to
do. Don’t look at your email, don’t look at Slack; look at your calendar, and
look at your task list.
One of those tasks, probably, is a daily reminder to “check your email”, but
that reminder is there more to remind you to only do it once than to prevent
you from forgetting.
I say “try” because this part is always going to be a challenge; while I
mentioned earlier that you don’t want to unthinkingly give in to
you also have to acknowledge that the reason it’s called a “cognitive bias” is
because it’s part of human cognition. There will always be a constant anxious
temptation to just check for new stuff; for those of us who have a
predisposition towards excessive
have it more than others.
We all need to make commitments in our daily lives. We need to do things for
other people. And when we make a commitment, we want to be telling the truth.
I want you to try to do all these things so you can be better at that. It’s
impossible to truthfully make a commitment to spend some time to perform some
task in the future if, realistically, you know that all your time in the future
will be consumed by whatever the top 3 highest-priority angry voicemails you
have on that day are.
Email is a challenging social problem, but I am tired of email, especially the
user interface of email applications, getting the blame for what is, at its
heart, a problem of interpersonal relations. It’s like noticing that you get a
lot of bills through the mail, and then blaming the state of your finances on
the colors of the paint in your apartment building’s mail room. Of course, the
UI of an email app can encourage good or bad habits, but Gmail gave us a
prominent “Archive” button a decade ago, and we still have all the same
terrible habits that were plaguing Outlook users
in the 90s.
Of course, there’s a lot more to “productivity” than just making a list of the
things you’re going to do. Some tools can really help you manage that list a
lot better. But all they can help you to do is to stop working on the wrong
things, and start working on the right ones. Actually being more productive,
in the sense of getting more units of work out of a day, is something you get
from keeping yourself healthy, happy, and
well-rested, not from an email filing
You can’t violate causality to put more hours into the day, and as a
frail and finite human being, there’s only so much work you can reasonably
squeeze in before you die.
The reason I care a lot about salvaging email specifically is that it remains
the best medium for communication that allows you to be in control of your own
time, and by extension, the best medium for allowing people to do creative
Asking someone to do something via SMS doesn’t scale; if you have hundreds of
unread texts there’s no way to put them in order, no way to classify them as
“finished” and “not finished”, so you need to keep it to the number of things
you can fit in
short term memory.
Not to mention the fact that text messaging is almost by definition an
interruption - by default, it causes a device in someone’s pocket to buzz.
Asking someone to do something in group chat, such as IRC or Slack, is
similarly time-dependent; if they are around, it becomes an interruption, and
if they’re not around, you have to keep asking and asking over and over again,
which makes it really inefficient for the asker (or the asker can use a
@highlight, and assume that Slack will send the recipient, guess what, an
Social media often comes up as another possible replacement for email, but its
sort order is even worse than “only the most recent and most frequently
repeated”. Messages are instead sorted by value to advertisers or likeliness
to increase ‘engagement’”, i.e. most likely to keep you looking at this social
media site rather than doing any real work.
For those of us who require long stretches of uninterrupted time to produce
something good – “creatives”, or whatever today’s awkward buzzword for
intersection of writers, programmers, graphic designers, illustrators, and so
on, is – we need an inbound task queue that we can have some level of control
over. Something that we can check at a time of our choosing, something that
we can apply filtering to in order to protect access to our attention,
something that maintains the chain of request/reply for reference when we have
to pick up a thread we’ve had to let go of for a while. Some way to be in
touch with our customers, our users, and our fans, without being constantly
interrupted. Because if we don’t give those who need to communicate with such
a tool, they’ll just blast
@everyone messages into our slack channels and
@mentions onto Twitter and texting us
Hey, got a minute? until we have to
quit everything to try and get some work done.
Questions about this post?
Go ahead and send me an email.
As always, any errors or bad ideas are certainly my own.
First of all, Merlin Mann, whose writing and
podcasting were the inspiration, direct or indirect, for many of my thoughts on
this subject; and who set a good example because he
wouldn’t have answered your email even back when his job was thinking about email.
Thanks also to David Reid for introducing me to Merlin’s work, as well as Alex
Gaynor, Tristan Seligmann, Donald Stufft, Cory Benfield, Piët Delport, Amber
Brown, and Ashwini Oruganti for feedback on drafts. Finally, thanks to reader
Rob for pointing out some minor errors in the originally published version.