To a better internet, aka RIP AIM
Social media is not a monolith.
Social media is everyone’s favorite internet to hate. Why? Social media promotes hate speech. Social media melts brains. Social media causes depression. Social media ruins democracy. And on and on and on.
But what we call “social media” is a blanket term websites and apps that purport to allow you to (a) build or find a network of people and (b) communicate and interact with that network. From the business side there’s also (c), allows the platform to sell advertising based on this activity. But as users, the “find and interact” isn’t fine-grained enough to support meaningful experiences. Saying Snap Chat is in the same category as LinkedIn is like saying Seventeen Magazine is in the same category as the Harvard Business Review. Sure, they are both publications, but the similarities mostly stop there.
Lumping them together into a singular category stymies our efforts to fix the problems we have by obfuscating the fact that these different platforms aren’t all the same, nor should they be. We would not expect Seventeen and HBR to make the same editorial decisions, and yet we seem to be demanding that all of “social media” do so. Users are coming to these platforms for different reasons, but ignoring their differentiation facilitates problematic business behaviors, bad business decisions, bad policy, and poor user experience.
We know users want to connect. What we’ve neglected to ask is why. By not asking this question, and not designing around the answers, we’ve built companies that still don’t really understand what their users are trying to do and are trying to fill those knowledge gaps with design decisions that fulfill business, not user, goals.
We should isolate, identify, and define the use cases that “social media” encompasses, and as we design and develop tools that utilize the internet and web technologies we should build for users’ needs, as fine-grained as we need to. We should challenge all of the assumptions we have about the space, because they don’t come from a place of understanding users. These are assumptions like: it has to be browser-based; it has to have “reactions”; there has to be discoverability; I have to be able to re-share; it has to be free; there must be ads; it has to be all-in-one; and so on and so forth.
Right now, we assume that all of these features must be present in order for a website or app to be successful. I believe this is false. Not every need I have as a user requires all or even any of these features. In fact, fulfillment of some of my needs even suffers by their inclusion.
Why does it need to be free?
Most of these assumptions come from our vast experience with a certain platform that has dominated the “social media” landscape to such an extent and for so long that almost every single one of their design decisions has become tablestakes. Facebook, in its pursuit to keep users within its ecosystem as much as possible, has blurred every line between every user need and every feature set, making it seem like each is inextricable from the other.
Facebook long ago decided that it can — nay, must — mash up all possible use cases into one contained ecosystem. The more tasks the platform can fulfill in one place the more data is available to advertisers and the more money they can charge for advertising. Any ecosystem convenience to users is simply a side effect. (I still wonder if any user ever has said, “it’s really important to me that my great Aunt be able to see that I just bought a desk chair from my neighbor on Marketplace”.)
In this race to keep users in your app for as long as possible — rather than solve the user need as well as you can — the losers are the users. Our experience, which should be design’s motivating force, has become an accidental consequence, an afterthought, of each of these decisions.
You can see this mentality play out with smaller platforms: it’s not enough to simply share a photo with a friend. You have to be able to react; comment; re-share; have a “disappearing” option. Soon, every platform is indistinguishable from every other, with the exception of the particular friend list you have on each one. (For me, my “friend” lists on each platform are a reflection of my social circle during the era in which it was popular, like physical address books that I can go back to someday if I need to get in touch.) I wonder, how much do each of those features really contribute to the richness of the human connections that we come for in the first place? Is my friendship served by being able to “like” or “re-share”? Or is it the platform and its advertisers who benefit from this capability?
Despite this, some platforms have sustained an experience that serves a more singular user need, and I would argue that those are generally the ones who are the subject of less public ire. If social media wants to turn down the heat on itself, platforms should start doing what their users want. This might change a lot about the landscape of the industry — users might have to pay for services, or companies might have to spin off “product” lines that don’t align with the core user proposition — but I’d argue it’s all for the better.
If we know what use cases social media platforms serve, policy makers would be able to target behavioral features of technology (e.g. discoverability). We’d be able to identify what feature sets are for users and which are not. We could build products that are accountable to users, and realign user needs with customer needs by making users customers again. All it takes is knowing what social media’s use cases are, and building within the constrained space. In a phrase, know our limits.
So, what are social media’s use cases?
Let’s rewind the clock. It’s 2004 (let’s say) and you’re about to graduate high school (let’s say). You spend a lot of your time on AIM talking to your friends when you should be doing your homework, and Facebook is the new thing where you can sign up for an account and become “friends” with your new college-roommate-to-be, the people on your floor, other people at your school who are equally as pretentious as you are. You probably had a MySpace or a Xanga or a LiveJournal back in middle school for like 5 minutes but soon realized that if you didn’t really enjoy reading your angsty peers’ ill-formed thoughts, they probably didn’t enjoy reading yours either. Asterisked exception for music and fanfic which seems to really have found a niche.
Sometimes you played Snood because it was the best game ever and honestly I just had to throw it in here for that reason. Snood is a software program that you downloaded, and not a browser-based game. So, for that matter is AIM. You might have been using a browser-based email client (I got my Gmail Beta invite in 2004?); your Yahoo or Hotmail was also via browser, and you obviously had accounts on all of them because that’s how we’ve always done it. Before, you might have had AOL, which if you recall was an entire self-contained ecosystem in a desktop app, in some ways a very closely related ancestor to what Facebook is now. Even earlier, in the late 90s, we searched with AltaVista on Netscape Navigator and checked our mail on a desktop e-mail client. These are cool historical artifacts but this observation actually has huge implications for the “social media” and internet economy as it exists today.
The point is, from all this, what I learned from being an internet native in middle school and high school is that there are three basic use cases for normal humans: doing things with people (AIM); doing things to people (MySpace); and doing things from people (AltaVista). This may sound weird, so let me explain.
—
Use case 1: talking with friends. This is the o.g. From ages 11 through high school, I spent a lot of time on AIM. AIM was basically safe — unless you went into a chat room and made dumb choices, no one could find you. In theory someone could randomly guess a screen name but they wouldn’t know it was you-you, and you could block them anyway. (Also, remember how asking for the screen name of someone you had a crush on was like the SCARIEST thing ever?) You could pretty much say whatever you wanted on AIM, and there was accountability because you more or less knew who you were talking to. There were also profiles and away messages, which allowed users to really express their individuality. AIM was the best thing ever and I mourn its passing daily.
Today’s solution for the problem AIM solved, which was talking with friends, is texting, WhatsApp, etc. You could consider SnapChat a more modern or multi-modal analogue. And yes, Facebook Messenger as a standalone app. What’s super important about this use case is it’s a two-way (or more) street. Conversations are by their nature consensual. One person saying words is not a conversation. In fact, one person saying words and another person “liking” those words is also not a conversation, as much as some platforms would like to convince us otherwise. In order to have a conversation, the interlocutor (i.e., your friend(s)) need to opt in to listening to what you have to say and, ideally, responding. We all remember the mortification of saying “sup” to someone you liked on AIM and they never responded, and that happened in private; today, that’s like posting a Twitter and getting no RTs. (See how internet I am?)
Today, nearly every platform has a DM feature that feels like an “add-on”. Having the feature can come in handy, but it also makes DMs subservient to the bigger, noisier ecosystem. (It also makes tracking our communications really hard — every day I see another tattoo artist announce they’re switching from Instagram DM to e-mail to handle bookings.) What if we took DMs out, and siloed them? Like they are in text, like they were on AIM.
I want to talk to my friends. But I don’t need to have this functionality available at all times on every app. The beauty of personal conversation is it’s private, it’s personal, and it’s real. We can’t be our authentic selves if we’re on display all the time, and we shouldn’t assume we want to be open to a personal conversation in a space that defaults to public visibility. On AIM or even in WhatsApp, you can only talk to people you know, or at least people to whom you provided your contact information. Contrast this with the ability that Facebook or Instagram or TikTok provides to make you visible to anyone who wants to find you. That is dangerous and uncomfortable in myriad ways that we didn’t have to address with AIM. It facilitates cyber-bullying, stalking, and harassment. It irreversibly blurs the line between what’s public and what’s private, prompting us to self-censor because we never know who might be listening.
As a use case, personal conversations need a space that’s private, safe, anonymous to the outside world and highly personalized or personalize-able. This breaks the business assumption that people can be targeted and ads can be sold. So, how do you make a business out of people talking with each other? You charge them to be there. Shocking, but there’s precedent for this. AIM was supported by AOL subscriptions, but there’s no reason you couldn’t pay a dollar or whatever to buy a piece of software that lets you connect to other friends on the network (App Stores, anyone?). You pay for your cell phone, your VoIP, and/or your landline (!), all of which are ways to communicate directly with people you know. Some of us even pay for email providers — after all, there’s no such thing as a free lunch.
tl;dr Use Case 1 is direct, private communication with individuals or groups. By definition, “social media” cannot fulfill this need because it obfuscates the boundary between public and private.
—
Use case 2a: “sharing life updates or baby pictures”, aka the modern Christmas card. The vintage (lol) analogy for this is probably MySpace/Xanga/LiveJournal, which has grown into blogging in general and to some extent all of the “content” creation platforms. The obvious difference between the digital and analogue iterations here, of course, is the ease of publication on digital. But there’s another, less obvious difference between updating everyone in your address book on the year’s personal highlights in your family, and blasting everyone you’ve ever interacted with with your every, ever-changing miniscule thought or rant.
That difference is the interaction model. With a Christmas card, you don’t really expect an interaction or a response. Your second cousin once removed doesn’t need to “like” or “comment” on your Christmas card; they could put it on their fridge, or just throw it in the recycling. You will literally never know, and that’s beautiful. You sent it to them because you want to keep them in the loop, not for your own personal validation. Even email follows this model to a large extent: you can send a newsy update to your listserv, and it doesn’t matter if you know who’s read it. You don’t expect many responses. You’ve done your duty by sending the update.
(You may notice I am trying to avoid the word “content” — even talking about “content” imposes a transactional or commercial relationship on the communication. People want to share their personal updates with people they care about, not sell “content” to their cousins. Most of us aren’t actually out here trying to build a brand.)
In contrast, when we publish digitally, whether it’s annually or daily, we have been taught to expect the people in our lives (or our “audience”) to interact: to react, to comment, to share. Hence, the proliferation of “stats”. Every blogging and e-newsletter platform, in addition to Twitter, Facebook, Instagram, TikTok, and more, tracks your “stats”, subversively training us to try to create “content” that will “engage” ever more and ever other users. These tools exist not because it matters to you, really, if your friend reads your post (you’ve just been trained to believe that this matters to you), but because it matters to the platform. These “stats” are the tools of marketers and ad salesmen, not of humans. Imagine if you had stats for your snail-mailed Christmas cards. I mean, can you imagine?!
I’ve noticed, as have we all, that certain types of pictures on Instagram net me more “likes”. Because the “goal” of Instagram is to maximize followers, likes, etc., I am incentivized to share pictures that get more of those things, not necessarily pictures that express me. This serves neither me — who wants to share updates about myself — nor my friends — who in all likelihood are probably not that interested in some mountain I just hiked, even if it was picturesque. This subverts my original goal as a user, which was to share updates about my life with the people in my life. It also subverts my friends’ goals, which were to get updates about my life, passively, when and how they wanted.
I would love to see a digital publishing tool that actually served the needs of the update provider (me) and respected the mental bandwidth of the update readers (my friends). Surprise, it might look like email — I have a friend who once sent semi-regular emails to a bcc-ed list with new chapters of his sci-fi novel. I loved this method because I could read or not read them, as I chose, without feeling guilty, but also feeling like I was part of the “inner circle”. It was like sci-fi chapters as Christmas cards. Friend’s goal: share his work, maybe get feedback. My goal: read his work, feel included, keep up to date with him. Any assumption of when or how or how frequently we interacted around these shared chapters was constrained to social construct, not embedded in the communication platform itself.
An even better internet tool for this that already exists are RSS feeds. RSS feeds allow readers to be notified when someone they’re interested in publishes an update, but there’s no feedback loop based on stats or engagement. There’s no ranking algorithm. It’s just, there is content or there isn’t content. If you want to read it, read it. If you don’t, delete it. It won’t get pushed back up to the top; you won’t see it again just because someone else read it. Websites fit this model well too, since the reader has every option to read or not read whatever has recently been added to the site. We need more internet like this.
We could distill much of this down into “consent of the reader”. Feeds with ranking or recommendation algorithms don’t ask for consent of the reader, only for more updates from creators. If you don’t want to read something — like if you scroll past it — it is likely to pop up again, especially if someone else “engaged” with it. Things you never even expressed interest in might pop up for you, based on these kinds of predictive models. Remember this for later: it’s a separate use case. You might feel this line-crossing instinctively in your gut, which tightens every time you see something “you didn’t ask for” pop up in a space you expect to be filled with things you did ask for (see: spam email; “you might like” recommendations; what Uncle Joe ate for breakfast).
The implication of designing for this use case — articulated more clearly, “an update I can share with friends AND reading only the updates I want to read” — is that direct interaction with an update becomes obsolete. The only things that happen to an update is it is sent, and it is received. Engagement is no longer monetizable, and no longer usable as a measure of interest which is then converted, ultimately, to a sales pitch to advertisers. It’s available for when you, the user, want it, but it’s not a critical portion of the platform experience.
If you send your life updates (hopefully infrequently) via email, and your email provider does not use your personal data to sell advertising which means you pay for your email provider but it also means they can build features that you, as an email sender and receiver, actually want (…like RSS feeds, hi Google Reader), how would that change the way email looked, felt, and operated? What if blogging or e-newsletter platforms for personal use lost their marketing stats features and turned into actual spaces you could go to see what was new with a friend? Maybe you have to pay some tiny fee to post an update, but this would also mean the platform isn’t incentivized to infinite-scroll or suggest new garbage, it’s just incentivized to show you what you came there to see.
tl;dr Use Case 2a is sending and reading updates. We’ve bungled this by obligating people to consume updates, and creating a toxic relationship between update creators and update consumers by treating every human relationship like a marketing interaction.
—
Use case 2b: your content is your business. Use case 2a described, I think there is definitely space for the “creator economy”, and we see this finally starting to separate itself from the ad-driven digital space of the layperson in the form of (parts of) Medium, Substack, Patreon, and OnlyFans. WordPress and others tried earlier by allowing monetization of blogs via advertising, but the direct-payment model seems more honest and wholesome. Plus, no one actually wants ads to show up in the middle of their hobby blog, especially ads they don’t control. Exhibit 1,000 in “ignoring users”: we were told that in order to be paid for our work, we had to accept that advertising was the only way.
I might put “influencers” in this category too, though that’s more of a subgroup of users than a platform. As an aside, I’d also argue the “influencer” economy is so robust and even extant only because social media was literally built to monetize the individual (see: stats).
Here, the content creator marketing tools, like number of clicks, opens, likes, shares, etc., make sense, because these are essentially business platforms. Your engaged readership is directly tied to your monetary income; you aren’t you, you are your business. You should, in a sense, write or create what your customers want, because that is how business works. Where the cognitive dissonance comes in, as I addressed above, is when platforms foist these transactional metrics on personal relationships.
Further, we don’t have the same scale of reader consent problems — seeing things we don’t want to see — in this space because content is already inaccessible if you haven’t opted in. Discoverability features likely contribute to some amount of user frustration, but by and large, I’m only going to see the things I’m already asking to (really, paying to) see.
An interesting middle ground between 2a and 2b are the “artist” social networks, like Ello, which are geared towards professional growth of artists. Content itself isn’t monetized as it is with OnlyFans, but presumably a working artist could use a following from this space to try to sell their work “offline”. I would suggest LinkedIn is similar: it’s a space with a specific purpose, which is professional networking. It suffers from a lot of the reader-consent problems from 2a, but provides a lot of useful discoverability functionalities (see Use Case 3), which users clearly need in order to execute their goal of finding new jobs.
Because this use case is more intentional and transparent in its connection to the commercial world, it’s easier to conceive of “pay to play”, whether it’s paying to have a membership to a professional network, or asking your customers to pay to have access to your creations. But it’s critical to take the “content-as-a-business” model out of the “personal-updates-from-friends” model: they are different use cases, with different rules, expectations, and assumptions.
tl;dr Use case 2b is for people trying to turn their content into a business. We’ve confused this with 2a by assuming people are trying to create marketable content. Separate “I want to market my content” from “I want to share my update” and we have a much healthier internet ecosystem, both for creators and consumers.
—
Use Case 3: finding something new. Sometimes, you do want to find new stuff. Maybe it’s an interest group, as in o.g. FB Groups and AIM chat rooms, or on subreddits or deep in Stack Overflow.
But a lot of times, we don’t want to find new stuff. This is one of the biggest problematic assumptions almost every SM platform currently makes. The space where we see the updates from the people we care about is littered with trash updates we don’t care about, people we don’t care about, and often people we don’t even know.
Discoverability is desirable from the consuming user’s standpoint in only a few cases: when looking for new ideas (like things to read or listen to, or answers to questions), when looking for new communities, and when actively engaging in commerce.
Discoverability — whether it’s a search capability or “suggestions” — is one of the cornerstones of user value of Reddit, StackOverflow, and other forums. These platforms thrive because and only because they are arithmetic and populous destinations for information, chatter, and expertise. They are — and this is critical — democratic, in the sense that answers are voted up or down, and the promotion of a message is clearly tied to its votes. This makes judging content straightforward, like a more-transparent Wikipedia. You came to find an answer and by george, you’ve found it. But most importantly, they also stick to their singular user need: ask questions, get answers. No bells, no whistles.
Though not its top-level user need, Reddit also presents a good model for finding new communities. It’s topical and diverse, and its searchability makes it easy to explore. Some blogging platforms (cough cough Tumblr) also seem to be good community-finding spaces, and I would chalk that up to a forum-esque culture of robust, and anonymous, commenting and cross-linking. MeetUp also provides a use-case focused, “find people in your area” product that is, if not always up to date, at least constrained to solving a single problem. There are a lot of mechanisms that can be used to help people find communities, but key to all of them is discoverability — and not confusing that with showing you things you already know.
Discoverability is also crucial in commercial contexts. Here I would include job searching (hence LinkedIn’s discoverability features being very important) as well as buying or selling things. Craigslist is a great model for how to buy and sell things in a non-obtrusive and highly consensual way on the internet. It meets very specific user needs: give me a place to buy a thing, and give me a place to sell a thing, and make it so that I see things near me. It doesn’t try to be more than it is; it’s not prompting you to “share your sale” with your friends or otherwise turn the interaction into anything more than it was. Contrast this with Facebook Marketplace, where buying and selling automatically becomes one of the constant stream of updates to your friends and acquaintances that I can almost universally guarantee you they do not care about. Not to say that Marketplace isn’t useful (gated access aside); but it doesn’t need to be a part of the broader “get/give updates from/to friends” experience.
Even in commercial contexts, discoverability often serves different user groups slightly differently. On blogging platforms, the writer wants to get discovered more than the reader wants to do the discovering. If I’m a creator on OnlyFans or Substack, I definitely want discoverability features because I want to increase my customer base. As a consumer of OnlyFans or Substack media, I may or may not want to explore new creators (after all, they’re just going to ask me for more money). However, it’s not an unreasonable assumption that if I’m already there to consume media I’ve paid for, I might well be interested in seeing what else is out there.
An analogy for this is suggestions from Spotify or Pandora or similar: you listen to X artist, you might like Y. Similarly, if you subscribe to B newsletter, you may also be interested in C. But this is very centered around what the user may be interested in: it’s a useful recommendation, it has context (e.g. music), and therefore is a useful discoverability tool. The platform itself doesn’t really benefit, their cut of subscription cost aside, from you liking or disliking their suggestion. They’re just a middle man for the transaction between creator and consumer, with little to no skin in the game.
When we see a discoverability mismatch is when we see mashups of commercial and personal use. Advertising on personal news feeds, for example: I am not going to the sosh meeds to discover a new underwear brand. I simply am not. However, advertisers want to believe that I am, and therefore they treat the space that in my mind is reserved for reading personal updates from friends (see use case 2a) as a space in which they are welcome to sell, because they assume I am buying.
Or, when we’re in our “personal” communication space and shown content or updates from people we don’t know, or do know but don’t really have much interest in. Once in a while these algorithms surface something useful — mostly on LinkedIn, so again a pseudo-commercial setting — but by and large there’s not much user value here. The value is data collection for advertising, and that’s serving a business goal, not a user goal.
Don’t get me wrong: lots of people discover new underwear brands this way. But I deeply question whether this is because they want to, or because this is just how it is. We should not assume that because someone wants to engage with their friends that they are also looking to discover new things to buy or new communities to join or even want to see what strangers have to say. I think this is one of the most fatal flaws of the modern social media platform. In some spaces, I do want to absorb new surprise content. In others, I absolutely do not. Control, control, control.
Rather than trying to force one user need on another and coercing coexistence, like showing me content from people I don’t follow in a space where I usually come to see updates from people I already know, we need to recognize these are distinct user needs, and design spaces to facilitate those experiences separately.
tl;dr Use case 4 is a tool for a person looking, and we’re not always looking.
Keep Them Separate
None of these things NEEDS the functionality of another one of these things to completely fulfill its users’ needs. Sure, having a DM feature on a general sharing platform can come in handy, but it’s not going to replace the primary mode of direct communication. Having your search engine provider ALSO be your email provider is just not necessary. AltaVista was fantastic and they didn’t try to make email; that’s what Yahoo and Hotmail were for. Not a single user was upset that AltaVista and Hotmail were different. In much the same way that we don’t put toilets in our kitchens or gardens in our bedrooms, it’s okay to have to go to different online spaces to carry out different digital tasks.
I wouldn’t mind one program for catching up with friends, another program for buying a new lawnmower, and another for finding and listening to great podcasts. In fact, I would prefer this, and this probably best represents how I try to shape my digital space. I also don’t need these three things to know all the same things about me. The lawnmower salesman doesn’t need to know what music I like; my friends don’t really need to know which lawnmower I have (although, they’re welcome to ask). This separation feels extraordinarily natural to me — I don’t have happy hours with my friends at the lawnmower store while we have our headphones in listening to music, so why do we try to force these things together in digital space?
Social media suffers from trying to be everything to everyone all the time. It’s all connected by the common thread of “information sharing with people”, but to date its design fails to ask and answer the simple question of “what are you sharing, and why?”. If we disentangle the myriad reasons people use the internet to connect, we can (re)build a robust and diverse internet economy that is safer, provides better user experiences, and facilitates real and novel innovation.
