Soundtrack:

By now, you have probably heard of Bluesky, the text-centric social media app that’s something of a Twitter clone, right down to being founded, and then unceremoniously abandoned by, Jack Dorsey himself.

Billionaires beat on, boats against the current, borne back ceaselessly to the past, etc.

You may also have some preconceptions around it specifically regarding politics, or only think of it in terms of how it’s “competing” with X or Threads, or whatever.

I would ask you to put those thoughts aside for a moment.

What I think is sorely lacking in the Internet-Adjacent Businesspeople space right now is an actual examination of the platform itself on its merits, and so that’s what I’m going to do here, sans culture war nonsense.

Before we get started, I should disclose that I’m a BlueSky user; if you want to see how bad a Poster I am, go right ahead, but I’m not kidding when I say it’s terrible and I don’t care. Take a look, y’all:

Table of Contents

• Back To The Feed-ture

• You Have To Start Nowhere

• BlueSky’s Killer Onboarding Feature

• CAFO (Combined Algorithmic Feed Optimization)

• They Changed Our Minds For Us

• The Open Web Blues (…ky)

• Is This Good For A Free and Open Web?

Back To The Feed-ture

The first core feature of Bluesky that we should talk about is the “non-algorithmic,” “chronological” feed.

If you’re not familiar with these concepts, which are actually just the rollback of technological developments to more or less the original state of social media, here’s the deal:

Long ago, social media feeds showed only content from people you had Followed or Connected with, or whatever the nomenclature was, and did so mostly in the chronological order in which the content was made.

This followed not only the logic of how humans think about Time and Events and Narrative, but also the predecessors to social media: web forums.

In case you’ve forgotten what that is or looks like, here is what is possibly the most culturally significant web forum of all time, still technically alive if long past its glory days:

If your quibble with that claim is that 4chan deserves the title, then here’s a fun fact you probably don’t know:

4chan only exists because Something Awful founder Rich “lowtax” Kyanka banned hentai on SA, one of many of his claims to fame and infamy but above all, influence over internet culture that persists to this day.

Okay, before we get back to chronological feeds becoming algorithmic feeds: here’s a flawed overview of this whole thing. It does have some key dates, such as FB going algorithmic in 2009 and a few others. Honestly, the naive over-optimism in this article (IMHO) is kind of a perfect setup for all this.

Now, why would anyone want to “go backwards” from algorithmic feeds that inject content from accounts you’re not connected to and rearrange the order in which everything is shown?

Because people hate it, a lot, according to their posts on social media platforms.

If you, too, are one of these people who is always saying “I’m sick of seeing random stuff I never asked for in my feed” and/or “I hate seeing posts from 3 weeks ago” than a non-algorithmic feed is your simplest path to greater sosh meed satisfaction.

Alternatively, if you’re REALLY into algorithmic feeds, not only is there an optional Bluesky generated one (called Discover), but there’s also the opportunity to code your own algorithmic feed through the API, which is pretty cool.

Here’s a little overview and commentary on it, with some good general Bluesky advice, and some observations and thoughts I generally concur with:

If people hate algorithmic feeds so much, though, how did they become the de facto way social feeds work today?

You Have To Start Nowhere

Algorithmic feeds (probably) started innocently enough, conceived as a feature added to help new users that didn’t yet have a list of accounts they’re following to provide them with relevant content.

Other approaches to seeding a follow list for new users have been tested extensively over the years, often times in combination with an algorithmic feed.

Importing contacts from email or other apps or a phone contact management system is common, but raises privacy and user trust concerns; plus, a given email that a user has in their contacts might not be the one linked to that other person’s social media and so they won’t get added.

This was how Facebook went about things early on, back when the key metric they identified for retention was that users needed to add 7 friends in 10 days. Of course, as I’ll discuss later-that was a VERY different Facebook than today’s product.

You can also present people with popular accounts, or “Genres” or “Topics,” to select from, which is what streaming services go with…before they swizzle that into a potent algorithmic cocktail.

This has historically been ineffective in two ways for follow-based social media apps:

1. The Topics tend to be vague and poorly constructed, being useless at worst, and simply low-impact fodder to feed an algorithm at best.

2. Suggesting the largest accounts to all new users has predictable snowballing rich-get-richer effect where once an account cracks this list, it gets ever more massive, and tends to stay at the top.

So there’s an argument that it’s genuinely helpful for new accounts, and people are even saying so deep into the anti-algofeed Bluesky discourse:

The Discover feed is not without controversy; some people hate it when their posts are put on the Discover feed, mostly because of the type of interactions they get from this feed vs. from their followers.

A clear theme is that whatever Bluesky does or doesn’t offer, it all often comes down to putting users back in a place where they are faced with the privilege and the burden of choice.

BlueSky’s Killer Onboarding Feature

Starter packs are user-created lists of accounts that can follow a theme or reflect what a user thinks is the best of a particular type of account, or even just be their own, personal Best of Bluesky.

Alongside other more standard features for accounts looking to expand their Following list, whether those accounts are new or old, is this highly functional and human curation centered feature that Bluesky users love.

You can follow all the users in a pack easily, or hand-select them from a list, and the packs are easy to post and send and share. It’s a feature that felt fully baked out of the gate.

CAFO (Combined Algorithmic Feed Optimization)

Okay, if the idea of algorithmic feeds probably started, and works best, as a tool for new users, why is it everywhere?

The short answer is that social media platforms argue it makes everything better, and the data shows it.

They say people consume more content, engage with more content, and end up following a more diverse range of people and pages and groups or whatever.

Here’s where all the familiar and valid criticisms of all social media rear their respective heads:

1. Is more content consumption good?

   • Are people enjoying the minute 45 as much as they were enjoying minute 5, or are they just stuck in an addictive loop?

   • When people consume more content, does the quality of content degrade along axes such as truth and negative emotional impact? Is the misinforming and dangerous slop so prevalent because the hogs are so hungry?

2. Is more engagement good? Is engagement even meaningful on many of these platforms anymore?

   • What kind of engagement comes out of the push for ever more of it? Are all the extreme hate and shock comments a product of an environment which rewards those and in turn encourages the creation of “rage bait” posts?

   • Is much of this “engagement” just a braindead behavior on the part of humans, and also coming from bots?

3. Are people being pushed to see content from a longer list of accounts in order to promote diversity of thought, or is it just to hedge the platform’s fortunes against a cadre of powerful large creators?

   • If it’s diversification at all costs, will it push users to a point where they’re getting information from so many places that it’s impossible for them to vet the authority and credibility of any of these sources?

   • Is this just to prevent creators from having any leverage at all?

Critics say that these algorithms make the platforms money and do pretty much all of the bad things I allude to above, and more; the platforms insist they’re at the mercy of a user base that will abandon them if they turn down the pressure on that delicious algo slop spigot.

Meta even ran a pretty elaborate study about it, although it raises at least as many questions as it answers, many of which are touched on in this Hacker News thread in which it’s being discussed.

It raises many of the standard “is Number X going up a good thing” but there’s something more important here: the alleged ultimate proof being that they ditched Meta and Instagram for TikTok.

They Changed Our Minds For Us

TikTok was born in an era when we had already become Algo Feed Zombies. It is, as we all know, little more than the pure white hot power of its recommendation algorithm.

So when Meta deprived them of the algorithmic feed that they had become used to, or even dependent on, to navigate social media, they jumped ship for the biggest thing that also offered an algorithmic feed.

The thing is that this only happened after Meta cooked our brains into binary swipe-or-stop switches that don’t even really process the media flying by our faces, while building their platforms into massive messes where our networks lack social cohesion and there’s a glut of content it’s impossible to navigate or curate.

They slowly boiled the frog and then when the frog somehow evolved to thrive at 212 degrees Farenheit, they turned the burner off and said “LOOK WE TOLD YOU BOILING WATER IS BEST” when it jumped a pot over into some broth that was still steaming.

They blew out your Facebook friend list, and then got some of those friends to post way too much, and when that wasn’t enough “content” anymore, they brought into the slop factories to show you literally unreal videos of actually impossible things that were often disgusting or confusing or scary or arousing or all of those at once and you wouldn’t know how to “find” anything you “wanted” to see anymore if you tried, and if you tried, the platform would bend itself in any and every direction but the one you wanted to go until you gave up.

Starting anew outside the boiling water is work. It’s work on the users’ part to aggressively reintroduce intention and choice into their social media usage. It’s work on the platform’s part to not build a horrifying mess that is impossible to sift through without an algorithmic feed.

Patience will be needed on the part of both parties.

The Open Web Blues (…ky)

Bluesky delivers vastly more traffic per follower, as per the analysis of many publishers of their own traffic, and that’s linked to Bluesky’s intentional efforts to do this.

They’re also making it easier for publishers to track this traffic, instead of the other way around, as every other major platform has tended to do.

This has progressed to such an extent that many organizations may have vastly smaller follower counts on Bluesky than on X or Reddit or Threads, and yet already get more traffic from Bluesky than those other sources.

This whole approach is summed up nicely in this WIRED interview with Bluesky CEO Jay Graber.

And looky there, it’s also designed on top of an open source protocol that would allow someone else to make their own BlueSky or something entirely different.

Is This Good For A Free and Open Web?

Yes! YES! 1000 times yes!

It allows you to control and carefully curate your own information environment. Even the tough sledding of getting new users to build a follower list is essentially community-led and human driven.

It drives traffic to places outside of Bluesky and is built on a protocol that allows people to build things that are interoperable with Bluesky and yet also their own separate things.

Bluesky lets you see what you want and be what you want on its own platform, and won’t stop you from taking the party somewhere else.

In Part 2, we’ll look at how Bluesky might stay this wonderful way, and how it might accomplish a key strategic pillar: making money.