Mastodawn

Martin Holland Jul 17, 2023

Twitter's main competitor #Threads is now a little more than a week old, and I had already indicated that some of the initial growth was definitely visible in our numbers. That's why I have prepared another update (this is the translation of https://social.heise.de@mho/110727882695234961):

I wasn't the only one to notice, that it got noticeably quieter on #Twitter after the introduction of #Threads (with the exception of Europe, yet it was noticeable here too).

For a few days, in our analytics, it looked like the visits from #Twitter would decrease so much, that #Mastodon could finally come out clearly in front. But quickly the effect has completely disappeared again: #Twitter as a whole is ahead of #Mastodon again. If you take out the effects I described earlier, they again remain on par.

Also, I still see that visits from #Mastodon's have decreased if you add up all our accounts on social.heise.de: So the end of automatic retweets (https://social.heise.de/@heiseonline/110371699219107096) has hurt us overall, @heiseautos, @heisec, @macandi & Co. reach significantly fewer people on their own.

That's a shame, given how it was before. But I can't say, that it's surprising.

Also, it is significant that of the 20 best performing articles on #Mastodon, 8 (!) were published in November and December, 5 in January/February and 6 in March/April: only one article published after April 4 made it into the top 20. The heyday of the recent #MastodonMigration was a while ago.

Things look a bit better when it comes to engagement: 6 of the 20 articles with the most boosts and favs appeared after April 1.

All in all, a rather mixed picture: #Mastodon obviously can't benefit much from the situation on #Twitter. But it also may be, that the recent growth of the #Fediverse may simply be taking place elsewhere and to a much lesser extent in the German-speaking world.

Of course, my numbers are only limited and @heiseonline has a special target group. But as long as there are no better numbers, I will keep an eye on it.

Martin Holland May 25, 2023

For weeks at least as many people have been coming to heise.de via our account @heiseonline on #Mastodon as via our account on #Twitter.

I have said this multiple times already, but this was in part based on an assumption, now I have another clear sign supporting this. That's because our account in Twitter was largely silent since Tuesday (no, it was probably our fault and not that of #ElonMusk). But that means, that no one came to our current articles via our own posts own #Twitter and instead only via others who shared them. The numbers dropped significantly and no matter how I look, what's left is about 30% of our normal numbers. Coincidentally, we already had such an outage on an earlier Wednesday with an identical result.

So if I want to compare how many visitors come only (!) via our accounts on #Mastodon and #Twitter, I have to subtract 30% from the sum of all visits via Twitter.

The following picture emerges (mind you, these are the visits only via @heiseonline and the Twitter account to messages in our newsticker; our brands @heisec etc. are not included): Visits via both our accounts run almost parallel and have been for months. This confirms once more what I have been writing in the past.

Absolute numbers remain quite low, though. But that's another story.

This Thread in German:
https://social.heise.de/@mho/110429490556083418

Martin Holland (@[email protected])

Über unseren Account @heiseonline auf #Mastodon kommen seit Wochen mindestens so viele Leute auf heise.de, wie über unseren Account auf #Twitter. Das hab ich ja schon ein paar mal erläutert, aber da war ein wichtiger Teil geschätzt. Jetzt habe ich eine neue Zahl, die diese Annahme klar untermauert. Und zwar war unser Account auf Twitter seit Dienstag und bis vorhin weitgehend verstummt (nein, es war wohl unsere Schuld und nicht die von #ElonMusk). #TwitterTakeover (Thread 🧵 1/3)

Heise Medien on Mastodon

Martin Holland May 12, 2023

Now that @heiseonline is about to make changes that are likely to have an impact on the number on visits via #Mastodon (deactivation of automatic boosts), it's time for my third summary:

https://social.heise.de/@mho/110354256288090157

To be honest, I don't see a clear trend overall: While the number of followers continues to rise and will soon crack 50,000 🥳 visits via #Mastodon have just stopped the downward trend. On the other hand, #Twitter has recently seen an even sharper decline.

If you only take the posts of @heiseonline directly, then the tendency is again clearer (Twitter falls more than Mastodon), but with the other accounts from us (@heisec, @heiseautos etc.) it is all over the place (of course with significantly lower absolute numbers): Sometimes #Twitter is in front, sometimes #Mastodon, sometimes one rises, sometimes it falls. Individual articles still have too great an influence on such small accounts.

When they have to find find their audience on their own now, I expect this to change to the disadvantage of Mastodon.

The number of interactions on #Mastodon, on the other hand, is continuously decreasing, which is of course not so nice. Especially against the background that since it posts mostly automatically, many more links are shared on Mastodon in total: So there are many more posts to interact with. Also the number of followers is increasing.

However, I must write that I can only count #Likes and #Boosts, it's possible that the picture would look different if I could count comments, too.

Also interesting the distribution of visits over time via both platforms: Via #Mastodon 74% of the visits come on the first day (on Twitter its just 61%). However, Twitter catches up on the second day, after which its 88% (Mastodon) vs. 82%. After four days visits over #Mastodon amount only to 7%, on Twitter it's 12%. Seems to show, that posts mostly won't go viral here, but it might have an influence that I only count posts via the posts from @heiseonline, but all visits via Twitter.

Maybe you can summarize it like this: #Mastodon is not really growing at the moment (at least in German-speaking countries), but it is not shrinking either. The usage remains stable.
With #Twitter, on the other hand, I would say that it looks like a decline again. There is currently a small #Twitter migration from influencers to #BlueSky. But it's still too early for anything definitive:

What do you think? What could I still look for, what would you like to know? Apart from the absolute numbers, of course 😉 .

(I analyzed more than 6800 news articles, that were shared on #Mastodon, more than 630 of them were liked plus boosted more than 50 times each. On #Twitter I still find only 4 articles with more than 50 such interactions.)

Martin Holland (@[email protected])

Angehängt: 1 Bild Jetzt da bei @heiseonline Änderungen anstehen, die sich auf die Zugriffszahlen auswirken dürften (Abschaltung der automatischen Boosts) ist es mal wieder Zeit für einen Zwischenstand: Ehrlich gesagt sehe ich insgesamt keinen klaren Trend: Während die Followerzahl weiter steigt und bald die 50.000 knackt 🥳 ist bei den Zugriffszahlen über #Mastodon lediglich der Abstiegstrend gestoppt. Andererseits ging es bei #Twitter zuletzt sogar stärker herunter. (🧵 1/5)

Heise Medien on Mastodon

Martin Holland Mar 28, 2023

I wrote my first summary of @heiseonline on #Mastodon almost two months ago: Time for an update.

One thing in advance, even if it has become noticeably quieter here in the #Fediverse, the numbers remain pretty stable: Visits via Mastodon are still slightly below that via #Twitter, but as a reminder, I can only compare visits generated via our accounts on #Mastodon with all visits over #Twitter. So the actual numbers are even closer together. (@heiseonline on Twitter has 246.000 Followers, on Mastodon it's 45.500)

On the first graph I have summed up the weekly entries via Mastodon vs. Twitter since the #Mastodon account went live at the beginning of November. The steady growth is partly due to the fact that only articles shared both on Mastodon and Twitter are included, and their number has increased. Overall, more visits come via Mastodon than via #Bing or #Xing.

The engagement (#boosts plus #likes, no #comments) with posts of @heiseonline stagnates, but remains significantly above that on Twitter. I find only four posts from heise-accounts on Twitter in this time that have been shared at least 50 times, on Mastodon it's a whole 125 (!). At the same time, nothing really goes #viral on #Mastodon, the upward swings of visits are significantly higher on Twitter.

A new initiative that shows noticeable effects is #Verpasstodon (something like "missed-o-don"): Under this hashtag @heiseonline re-shares on weekends the most clicked articles of the past week. And they get klicked noticeably again.
There is no criticism directed at the practice, the hashtag can be muted without missing anything else.

Despite such ideas, 89% of the visits via #Mastodon come on the first two days vs. 81% over Twitter.

A few more numbers: In total, @heiseonline has more than 122,000 interactions (#Boosts and #Likes, overview in the second image "Interaktionen"), per article that's an average of more than 26. About 4700 articles were shared, more than 500 of them have more than 50 interactions. More than 1000 each:
https://mastodon.social/@heiseonline/109550376533582612
https://mastodon.social/@heiseonline/109670205634915816

Follower numbers of @heiseonline grow very slowly, but the speed has increased since the Twitter account has a link to the Mastodon-account. @heiseonline is now the largest German-language media-account, slightly in front of @tazgetroete.
In the USA, there is a larger one with @ProPublica, but a few regional media are particularly active (#Newstodon).

The third picture shows visits via Mastodon and Twitter over the last 80 days.

Hope this is interesting for english-speaking accounts, too:
@fediversereport @[email protected] @[email protected] @atomicpoet @mmasnick @jeffjarvis @Gargron

Martin Holland Feb 9, 2023

After almost three months and until the move to social.heise, I managed the Mastodon-account of @heiseonline (on mastodon.social) and published more than 1600 posts there, it is now largely automatic. Time for some insights and a few things I learned. Maybe some english-speaking accounts are interested as well? @[email protected] for example..

If I hadn't known how often the shared links were clicked, I'm sure the motivation wouldn't have gone that far, but #Mastodon doesn't make that analysis easy (of course). You simply can't see who comes to a page via Mastodon. That's why I only shared links with a URL parameter that can be used to see that (of course not traceable to individuals). This showed exciting things:
Already since November, the shared articles received almost as many visits via the Mastodon-account alone as via Twitter in total. This suggests that the numbers are comparable, although the Mastodon account has grown from 500 to around 40,000 followers in this period. On Twitter, @heiseonline is followed by over 240,000 accounts. To be honest, #Twitter simply doesn't bring in many visits - contrary to what #ElonMusk, for example, seems to assume. We're talking about a few percent here.

I'll add a chart to prove this: Light blue are the entries via Twitter in total (!) and purple those exclusively (!) via the Mastodon-account to the articles shared there. The gap is not large. Both curves are rising slightly because more and more articles per day have landed on Mastodon. On Twitter, of course, many more articles were shared automatically from the official account in the beginning, but since a few days the number of shared posts is identical.

The engagement is also very revealing: While there was only one tweet from @heiseonline in this (admittedly still quite young) year that was shared more than 50 times, the Mastodon account comes to over 30 posts with more than 50 boosts each. An impressive example in the screenshots attached and that was neither about Mastodon nor Twitter 😉 If you can still trust Twitter's search, only four tweets from #heiseonline were commented at least 10 times this year, at @heiseonline there are much more comments.

The practice of re-sharing posts that have already been shared has been criticized from time to time. Because there are no #algorithms on Mastodon that resurface older posts, they are lost from view much more quickly. The fact that re-shared posts still attract interest has been seen in the numbers, sometimes "re-boosting" brought as many visits as the original post. If you don't want that, you can "Hide shared posts" on the profile page.

Because there is no search worthy of its name on #Mastodon, hashtags are really important. From the start, I have therefore tried to assign several per post, including #News (if it is a "news"). The results can't be measured, of course.

It's getting late, so thanks #DeepL 😀

Personal account on #Mastodon:	https://hannover.town/@fingolas
Work account on #Mastodon:	https://social.heise.de/@mho