Posted August 19, 2022 by Neo Weeb Lord X
This 8 pager is made on a a whim to test posting content to twitter on some schedule. Supposedly, that is a thing social media likes: schedules and consistency. Spoilers, there are some interesting stats differences, but did not yield anything solid.
I kept to the schedule of every Monday, Wednesday and Friday for 3 weeks, with the exception of one instance (more on that later).
I'm using the performance of Catching Up: Risky and the Pirate Master as comparison.
The only notable difference
Is Impressions.
I believe this stats work similar to itch.io. It is the number of time twitter show my tweet to somebody.
Posting everything at once results in a U shaped impression graph. The first and the last few(usually 2) tweets gets highest amount of impressions, the rest hangs around 20% to 50% of the highest impression.
This might be due to twitter display my latest tweet on my profile automatically. The last few pages are potentially double dipping in impression when some one clicks on my profile.
Posting things on a schedule produced impressions chart in a decline. With the first tweet still getting the most amount of impression, usually double of the next tweet.
On a glance this feels as expected, like sequels, you lose people between every release. Why wouldn't sequential pages in a comic work the same? Except, impression is not people's interest. Impression is when the algorithm decide somebody might be interested in something.
So why does twitter distribute impression more evenly when you stagger the content release? I believe this is due to the delay reply. Threads on twitter is just tweet rely to another tweet. If you tweet all at once, twitter algorithm probably disregard that. But if you stagger the reply, I bet twitter think that is a legit rely in some sense and boost the visibility of the tweet.
The reason why I think this way, lies in that one time I broke the schedule. In the second week, I skipped posting on Friday in favor of tweeting an unrelated gif on Thursday instead.
"fan art" of Cult of the lamb. I rushed this one out to the game's release. I like how it turn out, check it out if you haven't.
https://twitter.com/NeoWeebLordX/status/1557853984945098752
The extra gap produced by this interruption produced a significant drop in initial impression to the following tweet. The impression metrics did not start to recover (still much lower impressions than previous tweet so far) until I reply with the next page. I suspect that while the schedule does help in some way, bulk of the metrics is moved by the sub-sequence reply.
And now, why all of the above is probably meaningless
Like counts.
Like counts in my experience stops growing around 2 days. While impressions will continue to grow over time, though diminishing. I understand that only very few people will actually like a tweet, but, If my tweets are in fact still reaching new audience. Shouldn't the like count also increase? Even if only slightly.
The most extreme example of this I have, is a tweet of mine getting retweeted by さちなまRemtairy(レムテイリー)(@sachinama072). I believe that is the artist of Karryn's Prison( a very well made adult video game).
At the time of writing Remtairy have 13.8k followers. On the day of retweet, the impression number shot up to something like 2k. For comparison, the best I do in the first day of a tweet is around 100. The like count quickly grew to 25 within that day. The tweet's impressions is now at 4k and the like count hasn't increased at all.
This lead me to believe that the impression metrics is actually useless, because a lot of that might just be re-showing my tweets to the same people over and over again.
The experiment I took didn't really increase like counts. It stay around 0 to 10ish, which I feel is my usual range.
It is entirely possible that I am just too small time for the difference to be detectable, but I can't do anything about that.
Also I don't have access to historic data regarding my tweets, if something funky happen while I wasn't looking, I just wouldn't know.
In short
I did a thing that didn't really net me any useful information. I am a bit disappointed, but hey, if social media is that easy, everybody would have a million followers.
Twitter continues to be barely worth investing my time in. However, in terms of just general nsfw art/comic dumps, I haven't got a good alternative. Itch.io is usable for that, but it is best for bigger projects, rather than things that trickles in.
I shouldn't say that in such definite way, I haven't test posting comic updates to itch.io in a piece meal fashion. I just don't think people would be happy with the spam.
This has been babby's first head bonk against twitter algorithm, there are absolutely nothing useful here, but maybe you got less bored after reading it.
Peace.