Thanks for the explanations, but I’m not noticing a big “external links” penalty on my own tweets. Found some discussion of this penalty via Google, so it seems real but maybe not that “massive”? Also some of it dates to before Musk purchased Twitter. Can you point me to anything that says he increased the penalty by a lot?
Ah Musk actually published Twitter’s algorithms, confirming the penalty. Don’t see anyone else saying that he increased the penalty though.
BTW why do you “protect” your account (preventing non-followers from seeing your tweets)?
Ah Musk actually published Twitter’s algorithms, confirming the penalty. Don’t see anyone else saying that he increased the penalty though.
‘The algorithm’ is an emergent function of the entire ecosystem. I have no way of knowing what sort of downstream effects a tweak here or there would cause or the effects of post-Musk changes. I just know what I see: my tweets appear to have plummeted since Musk took over, particularly when I link to my new essays or documents etc.
If you want to do a more rigorous analysis, I export my Twitter analytics every few months (thank goodness Musk hasn’t disabled that to try to upsell people to the subscription—maybe he doesn’t know it’s there?) and could provide you my archives. (BTW, there is a moving window where you can only get the last few months, so if you think you will ever be interested in your Twitter traffic numbers, you need to start exporting them every 2-3 months now, or else the historical data will become inaccessible. I don’t know if you can restore access to old ones by signing up as an advertiser.) EDIT: I looked at the last full pre-Musk month and my last month, and I’ve lost ~75% of views/clicks/interactions, despite trying to use Twitter in the same way.
As for the ‘published’ algorithm, I semi-believe it is genuine (albeit doubtless incomplete) because Musk was embarrassed that it exposed how some parts of the new algorithm are manipulating Twitter to make Musk look more popular (confirming earlier reporting that Musk had ordered such changes after getting angry his views were dropping due to his crummy tweets), but that is also why it hasn’t been updated in almost half a year, apparently. God knows what the real thing is like by now...
Musk has now admitted his link penalty is not ‘merely’ a simple fixed penalty on the presence of a link or anything like that, but about as perverse as is possible:
To be clear, there is no explicit rule limiting the reach of links in posts.
The algorithm tries (not always successfully) to maximize user-seconds on 𝕏, so a link that causes people to cut short their time here will naturally get less exposure.
Best to post a text/image/video summary of what’s at the link for people to view and then decide if they want to click the link.
So, the higher-quality a link is, and the more people & time spent reading it, the more ‘the algorithm’ punishes it. The worse a link is, the shorter, more trivial, more clickbaity, the least worth reading, the less the algorithm punishes it and rewards it with virality. This explains a lot about Twitter these days.
(This also implies that it may be a bit hard to estimate ‘the’ link penalty, if the algorithm is doing anything to estimate the quality of a link so as to punish good ones more.)
Thanks for the explanations, but I’m not noticing a big “external links” penalty on my own tweets. Found some discussion of this penalty via Google, so it seems real but maybe not that “massive”? Also some of it dates to before Musk purchased Twitter. Can you point me to anything that says he increased the penalty by a lot?
Ah Musk actually published Twitter’s algorithms, confirming the penalty. Don’t see anyone else saying that he increased the penalty though.
BTW why do you “protect” your account (preventing non-followers from seeing your tweets)?
‘The algorithm’ is an emergent function of the entire ecosystem. I have no way of knowing what sort of downstream effects a tweak here or there would cause or the effects of post-Musk changes. I just know what I see: my tweets appear to have plummeted since Musk took over, particularly when I link to my new essays or documents etc.
If you want to do a more rigorous analysis, I export my Twitter analytics every few months (thank goodness Musk hasn’t disabled that to try to upsell people to the subscription—maybe he doesn’t know it’s there?) and could provide you my archives. (BTW, there is a moving window where you can only get the last few months, so if you think you will ever be interested in your Twitter traffic numbers, you need to start exporting them every 2-3 months now, or else the historical data will become inaccessible. I don’t know if you can restore access to old ones by signing up as an advertiser.) EDIT: I looked at the last full pre-Musk month and my last month, and I’ve lost ~75% of views/clicks/interactions, despite trying to use Twitter in the same way.
As for the ‘published’ algorithm, I semi-believe it is genuine (albeit doubtless incomplete) because Musk was embarrassed that it exposed how some parts of the new algorithm are manipulating Twitter to make Musk look more popular (confirming earlier reporting that Musk had ordered such changes after getting angry his views were dropping due to his crummy tweets), but that is also why it hasn’t been updated in almost half a year, apparently. God knows what the real thing is like by now...
Musk has now admitted his link penalty is not ‘merely’ a simple fixed penalty on the presence of a link or anything like that, but about as perverse as is possible:
So, the higher-quality a link is, and the more people & time spent reading it, the more ‘the algorithm’ punishes it. The worse a link is, the shorter, more trivial, more clickbaity, the least worth reading, the less the algorithm punishes it and rewards it with virality. This explains a lot about Twitter these days.
(This also implies that it may be a bit hard to estimate ‘the’ link penalty, if the algorithm is doing anything to estimate the quality of a link so as to punish good ones more.)