News for hackers. Stuff that matters.
Re:Doctorow: Social Quitting by dibt on Monday January 09, 2023 @01:06AM (#34307333)
I wouldn't trust Cory Doctorow's anecdotal evidence on Twitter usage. His Like/View ratio is very low, and overall engagement for someone with 486k followers is some of the worst I've seen.

Part of the issue is he spams his followers with retweets constantly, all day. People may not know they can turn off retweets for someone. They may just mute them, and continue to follow.

It's very much like Cory to write a long article about this when really what's happening is him and the people he follows are becoming more irrelevant.

23 likes / 8,722 views = 0.26 https://twitter.com/doctorow/status/1612122825288208387

44 likes / 9,384 views = 0.46 https://twitter.com/doctorow/status/1611993024879554560

193 likes / 15,400 views = 1.25 https://twitter.com/doctorow/status/1611976607845224451

44 likes / 29,300 views = 0.15 https://twitter.com/doctorow/status/1611858168774098945

162 likes / 48,500 views = 0.33 https://twitter.com/doctorow/status/1610282255057813504

I had to scroll a long time just to get those 4 because he's constantly tweeting.

1.0 is what you would want. Anything around 0.5 or less is really bad.

He hardly has any replies to his tweets. IMO, replies and quote tweets are what people are using twitter for. He also links off twitter too much.

SocialBlade stats: https://socialblade.com/twitter/user/doctorow/monthly

Re:Doctorow: Social Quitting by onion2k on Monday January 09, 2023 @02:30AM (#34307775)
Elon's latest tweet has a like to view ratio of 180k to 20.2m, or 0.009.
Re:Doctorow: Social Quitting by dibt on Monday January 09, 2023 @02:43AM (#34307839)
You picked the most extreme example of any account. I'd bet Elon gets a lot of hate-follows. They won't like his tweets. You can see many real people replying to his tweets, so you can tell he gets plenty of engagement.

For others without much engagement (retweets/replies), like Cory, all we have to go off of is likes, views, followers. I don't disagree that we can come up with a more accurate formula.

Also, I gave percentages, so it's more like 0.9.

Re:Doctorow: Social Quitting by onion2k on Monday January 09, 2023 @03:01AM (#34307919)
I'm pretty sure you could pick literally any account and get similar numbers. The point is that Cory isn't an outlier or unpopular or anything. All twitter users are like this.
Re:Doctorow: Social Quitting by dibt on Monday January 09, 2023 @03:29AM (#34308068)
> I'm pretty sure you could pick literally any account...

> All twitter users are like this.

That's the point. Cory isn't like all users. He has 480k followers.

The reality is very few of his followers engage with the tweets. He previously had engagement, he no longer has that engagement, and thinks Twitter is dying. It didn't help that he told his followers to go to mastodon.

I really think he doesn't play to the platform's strengths: succinct commentary. His whole career has been about long-form writing.

Re:Doctorow: Social Quitting by onion2k on Monday January 09, 2023 @03:47AM (#34308166)
So the argument here is that if you have few followers or loads of followers then the ratio of views to likes don't matter, or 1% is fine, but if you're in the middle with 480k it's very important and shows that you're out of touch if you only get 1%?

That doesn't seem particularly reasonable. If only 1% of users at any level are really engaging with the platform then its fair to say the platform has a problem.

Re:Doctorow: Social Quitting by dibt on Monday January 09, 2023 @04:38AM (#34308497)
> So the argument here is that if you have few followers or loads of followers then the ratio of views to likes don't matter

The ratio does matter. The person above used Elon Musk. The second most followed account on twitter, so it's not a great example, but does stick to the range I expect. I doubt his ratio would get below .05% often.

More (active) followers > more views > higher chance of likes

Twitter uses likes (& replies, rewteets, etc) as a signal to give it more visibility in your follower's feed. More visibility, means more likes. It's a feedback loop.

It makes sense when you consider there's more tweets being produced than time to consume them.

> but if you're in the middle with 480k it's very important and shows that you're out of touch if you only get 1%?

I've noticed the ratio rarely goes above 1% for any account. Someone popular, regardless of follower count, will consistently have between .05% and 1% like-to-views ratio. I've seen viral tweets get above above 1%, but not more than 2%.

Another wrinkle is that followers will turn on notifications for popular accounts, which shortcuts the algo.

If it DOES go below 0.05% frequently, it means they aren't regularly producing engaging tweets. It's more more noteworthy for a 480K follower account because they have so many chances to get views if only they had something worth viewing.

I don't know if you actually use Twitter. The feed isn't meant to be like an RSS feed. You are meant to see what is happening in the moment. It's fine if Cory treats it as an RSS feed, but most of his 480k followers aren't seeing his stuff. The few that do see it, don't think it's worth engaging with it, so it dies there.

Musk's ratio on that tweet is within the ranges I've noticed, but there are many other signals to pay attention to on his tweets that are missing from Cory's.

This ratio is stronger on private accounts, since views aren't diluted with non-followers. I see one with 773k followers frequently get a ratio above 1%.

> If only 1% of users at any level are really engaging with the platform then its fair to say the platform has a problem.

I think this is just the nature of the beast. I hardly ever like a tweet. I suspect the mechanics of the algo are involved. Like it might dampen popularity, or not promote, a tweet if the ratio goes above 1%.

Re:Doctorow: Social Quitting by dibt on Monday January 09, 2023 @11:22AM (#34312075)
> So the argument here is that if you have few followers or loads of followers then the ratio of views to likes don't matter

Ratio between 0.05% - 1% shows good engagement. I rarely see it above 1%. Those are 'viral' tweets. I've never seen 2% or more.

Ratio regularly dropping below 0.05% without lots of replies/retweets is bad. A low ratio could mean high views from bots (e.g. Musk) or other non-follower sources, so replies/retweets help to indicate there is healthy engagement. Cory doesn't have that.

> if you're in the middle with 480k it's very important and shows that you're out of touch if you only get 1%?

The 480k is his _potential_ reach.

You aren't entitled to reaching all 480k of your followers. They follow other people. Their feed has limited space.

Similar to how Youtubers tell people to subscribe AND enable notifications. Doing so shortcuts the algo. You can do the same on Twitter. If the small % of followers that see your tweets don't engage, the feedback loop of getting to more of your followers never kicks off.

This is obvious with well-liked accounts. More obvious when these types of accounts are private because the like/view ratio is a higher signal (ie no fly-by views).

> If only 1% of users at any level are really engaging with the platform then its fair to say the platform has a problem.

Not a problem. It's the nature of the beast. Only a small % engage. The likes create a feedback loop that gets more views, but rarely breaks above 1% likes/views ratio. You see the same limits to conversion rates in other areas (e.g. advertising or eCommerce sales).

Re:Doctorow: Social Quitting by scatters on Monday January 09, 2023 @02:46AM (#34307852)
That's 0.9. It's a percentage. 0.9 is not great, not terrible.
Re:Doctorow: Social Quitting by awb on Monday January 09, 2023 @02:52AM (#34307877)
I think your numbers are off by 100x or missing a %.

23 / 8,722 = .0026 or 0.26%

And by a desired ratio of 1, do you mean 1%?