3 ms·

One thing to note: Engagement will always be better on new platforms, because they a) have no bots and b) haven't been growth hacked to death, since the strate

One thing to note: Engagement will always be better on new platforms, because they a) have no bots and b) haven't been growth hacked to death, since the strategies that work haven't been figured out yet. These start at 0, then increase. When both of those trends grow strong enough, people start to leave in droves. I know I did. On X, I used it a lot, then left when the bad algo content era began. On Insta, the engagement bait was bad enough when I joined that it deterred me from ever seriously using it. I would expect these to become more of a problem over time, though I'm optimistic Bluesky can do better at beating them than other platforms have.

the_mitsuhiko 2y ago
Maybe. But at the time mastodon had a massive influx after the X acquisition and it never felt this alive. threads is massive and I don’t manage to get any engagement there either.
nasso_dev 2y ago
A major difference with Bluesky tho, is that users can choose their own "algorithm". Nobody is forced on Discover (Bluesky's "for you" algorithm). I wonder if this will play a role.
- tptacek 2y ago
  I was commenting on Twitter the other night just how bad Twitter's recommendations --- now hard to avoid, since everyone's funneled into the official web app --- have become; it's all random US political figures that don't intersect any of my interests. I likened it to 2003, when people fled Windows for the MacBook over issues like crapware (which: the X algorithm basically is). And: a lot of early MacBook adopters, most especially the ones that had been there for a long time, were insufferable! But we hit a point where one product was clearly better than the other, and you roll the tape forward and see how good the long-term prospects are for the niche product that just does what all its real users actually want it to. Hard to imagine why someone would go through so much trouble to acquire Twitter just to Ballmer-ize it. You usually have to have started and built up a product yourself to screw it up this badly.
  - idlewords 2y ago
    I found that some aggressive gardening of the 'For You' feed on Twitter made it borderline useful (By gardening I mean clicking the 'not interested in this post' option pretty aggressively). Eventually this seems to revert the recommendation algorithm to pre-Musk behavior. That said, this might be different on phones than on the web app. To your larger point, I think BlueSky is hobbled right now by the religious committment to decentralization, which the new arrivals don't care about at all. I would compare the situation to a world where people had to flee Windows for Linux—neither alternative is one that just works the way you'd want it to.
    - tptacek 2y ago
      What ever they're doing with decentralization isn't really legible to me. As much as I'll say nice things about Bsky, I'm not really rooting for it. As you saw: I left Twitter after Musk bought it and landed in the Fediverse, which I actually quite liked, not out of any religious interest in architecture so much as for the promise I thought (still sort of think) it held for restoring a blogosphere-type organization to the Internet (a world where most people would just sign up for whatever the Blogger.com of Mastodon was). Not enough people stayed in the Fediverse, and I found early Bsky culture, hm, difficult to mainline day-in day-out, so I went back to Twitter, and I'd be fine, regardless of ownership, if it would just be a good product that everyone wanted to use. So to me, Bsky just felt like Twitter with different owners; it still sort of does. Which: if you're looking for a less-shitty Twitter, to me, that's Bsky's offering today.
      - idlewords 2y ago
        I completely agree with you—the large exodus from Twitter means that both sites have the Twitter feel, so people can choose. My point about decentralization is that it imposes some weird technical limits on Bluesky for reasons that the normal users it is now attracting have zero interest in. For example, the list of who you block is public, and you can't quietly make someone stop following you by quickly blocking and unblocking them. I'm sure there are additional limitations, too. All of these stem from highly religious architectural choices that are of no interest to 99% of the current user base, who just want pre-Musk Twitter back. The obvious endpoint is what happened to email—there will be one central provider who can set arbitrary policy, and the decentralization part will just linger on as a series of annoying hoops that developers have to jump through to work with the de facto centralized site.
aussieguy1234 2y ago
I believe BlueSky, based on the way it's open source AT protocol works, would make it much harder to implement an AI/algorithmicly driven feed in the same way this happens on other platforms. And if it did happen, you could simply swap this Feed out for a different one.
- yen223 2y ago
  There's nothing preventing Bluesky, or someone else, from building an "algorithmic" feed on top of the AT protocol. That's what the Discover feed is, for all intents and purposes. What is harder is Bluesky forcing users into their flavour of an algorithmic feed, the same way e.g. X doesn't let you remove the Trending section, or how Reddit always shows you comments ordered by "Best". Bluesky users currently can choose the feed they want to use.
  - aussieguy1234 2y ago
    You have a point, but custom Feeds are one the main features of BlueSky. I have several Feeds for different interests, Twitter doesn't let you have these types of custom Feeds. I think it's unlikely they'll ever remove it. I can always ignore the Discover feed.
czhu12 2y ago
It’s almost like the opposite of the network effect moat that people used to think about social networks. At this point it feels like when these platforms hit a certain size, it’s time to move on to something smaller, without advertisers, less engagement boat, etc.
tptacek 2y ago
I don't think it's bots. I think it's that Twitter's owners have made it a product goal to pessimize this particular metric, and its engineers have succeeded. That's literally the logic ownership has used when explaining things like "links must be in replies, not standalone messages, or else they're penalized".
- yen223 2y ago
  A user clicking on a link takes the user away from the platform. If your goal is to maximise time spent on your platform, you'd be incentivised to suppress links as much as you can. That appears to be what's happening on X.
  - tptacek 2y ago
    Yes, that's what I'm saying. That "engagement" on Twitter isn't coming back, unless ownership changes the product to allow it. Bsky isn't going to grow out of this. Engagement is going where it's rewarded, and that's all there is to it.
    - rsynnott 2y ago
      > Bsky isn't going to grow out of this. It will if it’s a better user experience. Which, like, it _clearly_ is. Twitter may not want users clicking links, but users want links to click.
- llm_trw 2y ago
  Twitters owners used to be the current owners of blue sky. How gullible do you have to be to fall for the same trick by the same people twice? Edit: looks like I'm six months behind on the drama https://www.theguardian.com/technology/article/2024/may/07/jack-dorsey-quits-bluesky-board-urges-users-stay-elon-musk-x-twitter https://www.theguardian.com/technology/article/2024/may/07/j...
  - danielbln 2y ago
    Yep, the current CEO strikes me as a different breed form the Dorseys and Musks of this world. Time will tell.
olivierduval 2y ago
Actually, focusing on "engagement" is already showing that the game is lost and that the we're seing another Twitter clone... :-( Because "engagement" is about "money" and "ad" and "sponsor" and "business"... and not about anything close to "social" or "society" or "human". So, sooner or later (but most probably "sooner"), you'll see clickbait and other usual tricks to grow up "engagement"... and any interesting post will disappear So I won't even matter to create an account or have a look. I'll wait for the next one, or the next next one... hoping some people will understand that quality content is not measured by some kind of "engagement" EDITED: if you matter enough to downvote, at least explain WHY you think I'm wrong ;-)
- verdverm 2y ago
  1. Bluesky has stated they have no intention to add ads. Subscriptions are apparently coming soon 2. ATProto makes algorithms pluggable. Anyone can develop on, users can sub to those of their choosing, making switching easy should one become bad. It means there can be real feed competition without needing to move networks or apps. Same for moderation, individual choice and control is central to the design
  - olivierduval 2y ago
    The problem is not "ads" per se... it's engagement focus Ads or clickbait posts or filtering/ordering algorithms are just some consequences of engagement focus (also called "attention economy" because what matters is... well... attention... measured by "engagement") As long as focus is on "engagement" and not on "quality content", you'll have some Twitter/Facebook/Tiktok/... clone. Subscriptions COULD shift the focus on quality (because if people are convinced that you provide quality content, they might pay for it) but it requires (as shown for a long time) at least : - easy way to discover real quality content (a bit like Wikipedia in a way) - easy way to pay once to subscribe to different providers. That's the same old problem for the newspapers: people dont want to pay a yearly subscription - or even a newspaper issue - to read once a single article. They MAY pay a subscription if they can read any article in any newspaper (in a big bag of newspaper). Then this revenue has to be redistributed among the different newspapers...
Alex3917 2y ago
> Engagement will always be better on new platforms The primary effect is that communities with fewer users have stronger common bonds, the age of the platform isn't super relevant. That said, the engagement on Twitter has been terrible since almost the very beginning. I'm sure it was super magical for folks who were at SXSW that first year, but the conversion rate has been garbage ever since then. Whereas maybe 50% - 70% of people will open the average cold email, the percentage of people who will click a link on Twitter was maybe 1 in 1,000 even 10+ years ago.
- emseetech 2y ago
  I wonder if smaller, more active and tight-knit communities are healthier and preferable, but the combination of growth-at-all-costs, catering to influencers and fear of "bubbles" keeps those communities from becoming more common.
  - pixl97 2y ago
    Community is hard to maintain in the sense it either slowly dies off or experiences bouts of extreme growth even without the growth mentality. Think of a reverse slashdot effect where something posted on the community drives a bunch of users there quickly and either causes a rise in expense or a difficulty in moderation. If there is no growth you slowly lose users to entropy and the community fades.
    - emseetech 2y ago
      Feels like there should be ways to mitigate this, like having a waiting list and booting users for inactivity after a certain amount of time. That way the community stays active but doesn't get flooded.
meiraleal 2y ago
Changing platforms from time to time is beneficial then

the_mitsuhiko 2y ago

Maybe. But at the time mastodon had a massive influx after the X acquisition and it never felt this alive. threads is massive and I don’t manage to get any engagement there either.

nasso_dev 2y ago

A major difference with Bluesky tho, is that users can choose their own "algorithm". Nobody is forced on Discover (Bluesky's "for you" algorithm). I wonder if this will play a role.

tptacek 2y ago
I was commenting on Twitter the other night just how bad Twitter's recommendations --- now hard to avoid, since everyone's funneled into the official web app --- have become; it's all random US political figures that don't intersect any of my interests. I likened it to 2003, when people fled Windows for the MacBook over issues like crapware (which: the X algorithm basically is). And: a lot of early MacBook adopters, most especially the ones that had been there for a long time, were insufferable! But we hit a point where one product was clearly better than the other, and you roll the tape forward and see how good the long-term prospects are for the niche product that just does what all its real users actually want it to. Hard to imagine why someone would go through so much trouble to acquire Twitter just to Ballmer-ize it. You usually have to have started and built up a product yourself to screw it up this badly.
- idlewords 2y ago
  I found that some aggressive gardening of the 'For You' feed on Twitter made it borderline useful (By gardening I mean clicking the 'not interested in this post' option pretty aggressively). Eventually this seems to revert the recommendation algorithm to pre-Musk behavior. That said, this might be different on phones than on the web app. To your larger point, I think BlueSky is hobbled right now by the religious committment to decentralization, which the new arrivals don't care about at all. I would compare the situation to a world where people had to flee Windows for Linux—neither alternative is one that just works the way you'd want it to.
  - tptacek 2y ago
    What ever they're doing with decentralization isn't really legible to me. As much as I'll say nice things about Bsky, I'm not really rooting for it. As you saw: I left Twitter after Musk bought it and landed in the Fediverse, which I actually quite liked, not out of any religious interest in architecture so much as for the promise I thought (still sort of think) it held for restoring a blogosphere-type organization to the Internet (a world where most people would just sign up for whatever the Blogger.com of Mastodon was). Not enough people stayed in the Fediverse, and I found early Bsky culture, hm, difficult to mainline day-in day-out, so I went back to Twitter, and I'd be fine, regardless of ownership, if it would just be a good product that everyone wanted to use. So to me, Bsky just felt like Twitter with different owners; it still sort of does. Which: if you're looking for a less-shitty Twitter, to me, that's Bsky's offering today.
    - idlewords 2y ago
      I completely agree with you—the large exodus from Twitter means that both sites have the Twitter feel, so people can choose. My point about decentralization is that it imposes some weird technical limits on Bluesky for reasons that the normal users it is now attracting have zero interest in. For example, the list of who you block is public, and you can't quietly make someone stop following you by quickly blocking and unblocking them. I'm sure there are additional limitations, too. All of these stem from highly religious architectural choices that are of no interest to 99% of the current user base, who just want pre-Musk Twitter back. The obvious endpoint is what happened to email—there will be one central provider who can set arbitrary policy, and the decentralization part will just linger on as a series of annoying hoops that developers have to jump through to work with the de facto centralized site.

aussieguy1234 2y ago

I believe BlueSky, based on the way it's open source AT protocol works, would make it much harder to implement an AI/algorithmicly driven feed in the same way this happens on other platforms. And if it did happen, you could simply swap this Feed out for a different one.

yen223 2y ago
There's nothing preventing Bluesky, or someone else, from building an "algorithmic" feed on top of the AT protocol. That's what the Discover feed is, for all intents and purposes. What is harder is Bluesky forcing users into their flavour of an algorithmic feed, the same way e.g. X doesn't let you remove the Trending section, or how Reddit always shows you comments ordered by "Best". Bluesky users currently can choose the feed they want to use.
- aussieguy1234 2y ago
  You have a point, but custom Feeds are one the main features of BlueSky. I have several Feeds for different interests, Twitter doesn't let you have these types of custom Feeds. I think it's unlikely they'll ever remove it. I can always ignore the Discover feed.

czhu12 2y ago

It’s almost like the opposite of the network effect moat that people used to think about social networks. At this point it feels like when these platforms hit a certain size, it’s time to move on to something smaller, without advertisers, less engagement boat, etc.

tptacek 2y ago

I don't think it's bots. I think it's that Twitter's owners have made it a product goal to pessimize this particular metric, and its engineers have succeeded. That's literally the logic ownership has used when explaining things like "links must be in replies, not standalone messages, or else they're penalized".

yen223 2y ago
A user clicking on a link takes the user away from the platform. If your goal is to maximise time spent on your platform, you'd be incentivised to suppress links as much as you can. That appears to be what's happening on X.
- tptacek 2y ago
  Yes, that's what I'm saying. That "engagement" on Twitter isn't coming back, unless ownership changes the product to allow it. Bsky isn't going to grow out of this. Engagement is going where it's rewarded, and that's all there is to it.
  - rsynnott 2y ago
    > Bsky isn't going to grow out of this. It will if it’s a better user experience. Which, like, it _clearly_ is. Twitter may not want users clicking links, but users want links to click.
llm_trw 2y ago
Twitters owners used to be the current owners of blue sky. How gullible do you have to be to fall for the same trick by the same people twice? Edit: looks like I'm six months behind on the drama https://www.theguardian.com/technology/article/2024/may/07/jack-dorsey-quits-bluesky-board-urges-users-stay-elon-musk-x-twitter https://www.theguardian.com/technology/article/2024/may/07/j...
- danielbln 2y ago
  Yep, the current CEO strikes me as a different breed form the Dorseys and Musks of this world. Time will tell.

olivierduval 2y ago

Actually, focusing on "engagement" is already showing that the game is lost and that the we're seing another Twitter clone... :-( Because "engagement" is about "money" and "ad" and "sponsor" and "business"... and not about anything close to "social" or "society" or "human". So, sooner or later (but most probably "sooner"), you'll see clickbait and other usual tricks to grow up "engagement"... and any interesting post will disappear So I won't even matter to create an account or have a look. I'll wait for the next one, or the next next one... hoping some people will understand that quality content is not measured by some kind of "engagement" EDITED: if you matter enough to downvote, at least explain WHY you think I'm wrong ;-)

verdverm 2y ago
1. Bluesky has stated they have no intention to add ads. Subscriptions are apparently coming soon 2. ATProto makes algorithms pluggable. Anyone can develop on, users can sub to those of their choosing, making switching easy should one become bad. It means there can be real feed competition without needing to move networks or apps. Same for moderation, individual choice and control is central to the design
- olivierduval 2y ago
  The problem is not "ads" per se... it's engagement focus Ads or clickbait posts or filtering/ordering algorithms are just some consequences of engagement focus (also called "attention economy" because what matters is... well... attention... measured by "engagement") As long as focus is on "engagement" and not on "quality content", you'll have some Twitter/Facebook/Tiktok/... clone. Subscriptions COULD shift the focus on quality (because if people are convinced that you provide quality content, they might pay for it) but it requires (as shown for a long time) at least : - easy way to discover real quality content (a bit like Wikipedia in a way) - easy way to pay once to subscribe to different providers. That's the same old problem for the newspapers: people dont want to pay a yearly subscription - or even a newspaper issue - to read once a single article. They MAY pay a subscription if they can read any article in any newspaper (in a big bag of newspaper). Then this revenue has to be redistributed among the different newspapers...

Alex3917 2y ago

> Engagement will always be better on new platforms The primary effect is that communities with fewer users have stronger common bonds, the age of the platform isn't super relevant. That said, the engagement on Twitter has been terrible since almost the very beginning. I'm sure it was super magical for folks who were at SXSW that first year, but the conversion rate has been garbage ever since then. Whereas maybe 50% - 70% of people will open the average cold email, the percentage of people who will click a link on Twitter was maybe 1 in 1,000 even 10+ years ago.

emseetech 2y ago
I wonder if smaller, more active and tight-knit communities are healthier and preferable, but the combination of growth-at-all-costs, catering to influencers and fear of "bubbles" keeps those communities from becoming more common.
- pixl97 2y ago
  Community is hard to maintain in the sense it either slowly dies off or experiences bouts of extreme growth even without the growth mentality. Think of a reverse slashdot effect where something posted on the community drives a bunch of users there quickly and either causes a rise in expense or a difficulty in moderation. If there is no growth you slowly lose users to entropy and the community fades.
  - emseetech 2y ago
    Feels like there should be ways to mitigate this, like having a waiting list and booting users for inactivity after a certain amount of time. That way the community stays active but doesn't get flooded.

meiraleal 2y ago

Changing platforms from time to time is beneficial then