My oldest brother is just like me, but several years behind.
For instance: I was on reddit as early as 2007, and watched the rise and fall of reddit atheism in real time. Years later, I heard him parroting the same talking points I had already heard beaten to death (in both directions), and all I could say was "mmm hmm."
So when he has problems he wants to solve, I've often been through an entire cycle of thought, and have some sort of easy solution for him that took me literal years to arrive at. The difficulty, for me, is trying to construct the right argument to help him see the path forward without having to make all the same mistakes, in the same order, as I did.
So when he asked me how to deal with organizing his email, I told him not to bother. He was a bit surprised, but as he heard my justification he seemed to accept it.
The fact of the matter is, in today's world, email is a lost cause. But I don't think we should give up, not yet; systems are for humans, and if the system has been taken over, it is incumbent on us to take back control. But it's not going to be easy.
How to sort email
I have long been used to the idea that no one else uses products like I do. I resent, with a fire others usually reserve for criminals, the dark pattern so often implemented by websites that a) require a signup, and b) don't give you the option to opt-out of emails during the signup flow.
As far as I can tell, most people just sign up and then weather the flood of emails that fill their inbox, drowning signal in an endless wash of noise, making it nearly impossible to see messages that actually affect their lives. My wife does this: her inbox is full of promotions and newsletters, and when she needs to get something done, she wades into the mess and finds what she needs. From my own observations, I believe her habits resemble those of the typical American.
My immediate family, by contrast, trends toward atypical. To be certain, there's a continuum of extremes, from my mother, to me, to my oldest brother, to the rest of my family... my mother is an absolute radical on matters like these; my father is much more sanguine. But just having three people who care about the sheer volume and quality of email we receive makes us, I think, exceptional outliers.
Personally, I went through an entire narrative arc when it came to handling email. In the early days (back in 2005 or so, when the promise of "unlimited email!" was still new), I read newsletters and sales flyers willingly. As time went on and the volume of email increased, my relationship soured; I started unsubscribing from things I wasn't reading anymore, then deliberately deleting archived messages, then aggressively unsubscribing the instant anyone sent me anything at all, and now... I sit somewhere in the middle, understanding the forces that have shaped the world, and dealing with them, but still believing in my heart-of-hearts that we can do better.
What does this look like, practically?
Well, to start, I still unsubscribe from newsletters and sales flyers, but I am close to enough marketing people that I understand why they exist. If an email address is required for signup, email is the one way you can be sure you can engage with users. And, for marketers, their product is The Most Important Thing in the World™, so being able to tell users about a new feature that might make them more excited about using the product, or willing to come back, is invaluable. But I don't have time for that. You get one chance to impress me, and if you muck it up, I'm out.
What's left mostly fits into a broad class that can be called "transactional emails." These are things like statement notifications, invoices, and bills from people I pay money to; account registration and update emails; "legal" documents (construed broadly; things like jury duty, marriage, immigration, and the like); and so forth. These, I bucket into coarse folders ("labels", in the gmail parlance) using a one-pass system: bills to Bills, invoices to Invoices, and so on. It's awful and incomplete, and nearly impossible to automate, but it mostly works for me.
After these two passes, the residue falls into two categories: humans; and messages that the upstream senders view as transactional, but that I view as spam. The latter get deleted with prejudice. The former get replies.
(As an aside, I am old enough to remember when by far the most common emails you received were ones you cared intensely about. Back in 2000, my university had sandboxed email-only terminals scattered across campus, and I still remember the thrill of logging in and seeing one (1) unread message, maybe from a friend, maybe from a girlfriend, containing who-knows-what excitement inside. I fear those days are gone and will never come back.)
Grappling with complexity
Back to my brother, then. Why do I sort at all? That is, if I told him that the Problem of Email is insurmountable, why do I, hypocrite that I am, even do a single pass?
There are a couple reasons, some bad, some good. The first and most obvious one is habit. While attempting to find a workable solution for my own email, I tried many things, and after I had poured sufficient effort into failed experiments, I was left with something that still kind of worked, or at least pointed toward something that worked. Thus, I maintain it.
Further, though, my system allows me to answer certain questions with ease. For instance, "which recruiters have reached out to me in the past six months?" Or, "what are all the websites I have accounts on, so I can move their login information into my password manager?"
For questions like these, where I don't always know in advance that I will need to answer the question at all, a one-pass triage gets me to, and keeps me close to, the ability to answer these quickly.
Right now, you might be asking, why not simply use gmail's search function to find those emails at the moment you need them? In response, I will say to you: can you easily construct a single search that will surface every account you have ever signed up for? What format, what common trait, do those emails have that you can leverage to find them all, without having some intelligent agent (e.g. you), appending metadata to them?
Even things that should be simple, like bank statements, are surprisingly hard to re-locate over time. First, the format of the emails change. Then their content. Then the name of the sender, then the sender's address, then the reply-to address. Sometimes the statement is attached; sometimes not.
I have had a relationship with the same bank for over 15 years, and I have had the same email address that entire time, and digging through those emails reveals archaeological strata to rival the excavation of Troy. There is no one search that can surface all of these, and anything that comes close also surfaces things like those messages "the upstream sender views as transactional," reminding me that my bank has an app.
All valuable systems become adversarial
"Why do you rob banks?"
"Because that's where the money is"
– Willie Sutton (apocryphal)
Email has become useless for a lot of reasons, but the biggest one is that everyone has it, and some of those people have money, and other people who use it want money.
There are others, of course. It was a minimal system, originally designed for trusted networks of trustworthy people, that has had endless layers of fixes grafted on to try to preserve some semblance of value. POP, then IMAP (both horrid protocols, for different reasons). Anti-spam filtering (using blocklists, allowlists, IP bans, bayesian filtering...). Encryption (PGP and GPG). Sender verification (DKIM and DMARC). And on and on.
It has gotten so bad that, if you want to run your own email server, and actually have your messages read by others, you have two choices: spend months or years of your life learning about 50 years of layered knowledge, and even then, never being sure that people will be able to read your messages, or use one of a small handful of email providers. Even then, using a custom domain on top of one of these providers is not enough to get through many spam filters; you must, at minimum, implement sender identify verification per-domain.
And even that's not enough. I have, on multiple occasions, seen gmail itself flag google's own emails as potential phishing.
If the system has become so adversarial that Google cannot even tell when an email is legit, what the hell are the rest of us supposed to do?
Surrendering the medium
We as a society (and we as an industry) have ceded far too much ground to adversarial actors.
Spam isn't the only problem, though. The last major innovation in email was gmail, 20 years ago; almost everything done in the space since then has been to a) copy gmail, b) add another layer of duct tape on the ecosystem, or c) add minor UX improvements on top of gmail-like systems (see: Superhuman).
What is email for? In many cases, it's the greatest common factor between people, the medium you can rely on everyone having access to. (In developing markets, this is less true; SMS and WeChat and WhatsApp predominate, and when users must create an email address, their inboxes often go unmonitored.) If you have someone's email address, you can bootstrap your way to any other form of communication you want; in many contexts, it's like the BIOS of the social internet.
(In its heyday, Facebook used to serve this purpose for me; with no more than a person's name, I could track them down on Facebook and use that to move to a more useful communication system. But I don't use Facebook anymore, and neither do most people I know.)
Interpreted through the lens of POSIWID, email is for random people to try to urge you to give them money, either legitimately or illegitimately. Indeed, it is how your bank and power company let you know your bill is due; but it's also how illicit boiler room spammers try to get you to buy dodgy medications. And outside some glacially slow changes (remember Google Inbox?), there hasn't been any innovation in almost 20 years to suggest otherwise; widely-used protocols do not steer quickly.
Indeed, the proliferation of communications platforms designed to work around email's shortcomings tells us much about what we want from email. For work communications, where we used to use email, we use Teams and Slack. For social communications, we use Discord and one of the many chat apps (Line in Japan, Kakao in Korea, WeChat in China; WhatsApp, iMessage, Signal, Telegram; Facebook/Instagram/Twitter DMs; whatever Google messaging platform happens to be working that day). For apps we allow in our lives, there are push notifications.
For what remains, there is and remains only email.
Moving away from email didn't save us
All of these platforms and providers are missing features. Google has no stability. DMs are sandboxed, and become inaccessible when users get banned or leave the platform. Slack and Teams cause as many or more problems then they solve: they supplant a small volume of email communications with an even larger and more frustrating volume of mixed-intent messages. Things people wouldn't dream of putting in an email, at any time in history ("how was your weekend?") become entire channels on work chat.
The systems we have presently divide communications into a taxonomy of uses. I will send this message to my friend via iMessage, thinks the user, because it has better link previews.
The calculus the user undertakes is something like the following:
- what is the subject I want to talk about? (Mostly personal matters, work, or social.)
- who is the intended audience? (My coworkers in their function as my coworkers; my anonymous followers; my doctor; this friend)
- what sort of interaction to I want to permit? (social media "engagement"; one-to-many replies; one-on-one conversation; fire-and-forget)
- what features of the message do I care about? (encryption; link previews; emoji reactions; ease of use)
In most cases, this determination happens in the blink of an eye.
There are times where our interaction models fall apart: sending the same meme to multiple recipients is unduly difficult (especially across apps); using the "wrong" medium or the "wrong" chat for someone with whom you share multiple channels can cause chat fragmentation; etc. But as the various channels converge towards feature-parity, "right" and "wrong" uses sublime away. The issues that remain exist largely at the edges. (How do I find one particular message that I know I wrote on one of six platforms during a particular six month period?)
Once upon a time, there were chat programs that unified chats across providers. From the late 90s into the early 2000s, savvy users would install desktop apps like Adium for Mac or Trillian for Windows, and use them to sign into multiple accounts all at once. (There are apps attempting to re-create this functionality today.) If you met someone who only used Yahoo! Messenger, you would simply go to yahoo.com, create a free account, then add that account to your multiplexing client. Easy as pie.
Facebook Messenger and early versions of Google Chat implemented the cross-platform protocol XMPP, which was supported by the multiplexing chat apps. But as these platforms increased in popularity and their owners realized what a relative advantage walling the garden could give them, they either broke compatibility or simply closed the gates. Further, AIM, which at one time was the de facto default messaging client, shut down completely. (Others, too, have come and gone.)
So we have actually regressed on this front; where it was once possible to just fire off a message at a person, users now have to think about how they want to interact with them. And, at the extreme, there are people who will cut others out of their lives for want of a compatible messaging client.
Don't invite green bubbles to the group chat.
I only have an instagram account to talk to her.
The proliferation in communications mediums has not made communications easier, but it has helped to isolate the emotional valence of each platform from the stresses of the others.
A hateful, writhing mass
Nothing has replaced email; but we have supplanted parts of it through the apps, and increased the volume of messages that people have to deal with by several orders of magnitude. In fact, the social uses of email have dried up so much that email itself is seen by most as a painful obligation:
why would i check my emails, that’s where the emails are
But if email were only pain, and not a place for things we know we have to deal with, people would simply not use it. (At least, I hope they wouldn't. With some people, you never know.)
People use email because it is a universal pipe.
Similar to XMPP, Simple Mail Transfer Protocol (SMTP) is the backbone of the email system; if you use a third-party email client, it's what your client uses to send email to your email provider's servers, and what your provider's servers use to send your email to the next server in the chain, and so on. But SMTP is a back-haul technology; an implementation detail. Outside setting up an email client (and I suspect most users don't even do this anymore, since setting up the most common providers is no more difficult than entering a username and password), users don't particularly care how their message gets to the recipient, only that it does.
But since email is a universal pipe, it can be used for literally anything, and that's where the big issues start. In fact, there's an illustrative parallel between Slack/Teams and email here: the same formlessness that makes email so anxiety-inducing also afflicts Slack and Teams communication. At any moment, information can land in your lap that requires you to make a decision (set of decisions...); if I have an issue, the quickest way to deal with it is to make it everybody's issue.
Similarly, two of the most important learnings about emailing effectively are that the person on the other end probably so hates dealing with email, that you should not write anything important past the third sentence, and you should not ask for more than one thing in a single email.
The recipient views it as part of their job to deal with email, but they will do so in the least painful way possible.
Can we eliminate the pain?
We sort of, kind of, know the ontology of features that users are looking for in communications tools. It's not something that anyone has formalized (that I know of), but like Justice Potter Stewart, users know it when they see it.
Slack and Teams and email all offer tools for managing the noise (mute, leave; filter, unsubscribe), but users still struggle. They are afraid of missing important information. They don't feel empowered to, or capable of, using the tools effectively. They have trouble re-surfacing information when they need it. And so on.
In general, users under-learn and under-use the features that are available to them, for a great many reasons. Some are good and some are bad, but in general, those reasons make sense from the user's point-of-view. (They don't feel like it's worth investing in this system, they don't trust the system, they don't believe that investing in the system will make their life easier, they don't even know that it's possible for things to be better, to name a few.)
I believe that there is a way to reduce the pain of email, and indeed, of all person-to-person communication substantially, but indeed, it will require an almost visionary approach to get there.