[steveklabnik]

As someone who spends a bunch of time paying attention to the YouTube meta, "consistency" is a virtue that's been professed for a long time, but YouTube has been pretty clear that it's not a factor in how recommendations work.

I still think it's decent advice for new YouTubers, because you only get better at a thing by doing it a lot, and over time. The value of consistency isn't in the algorithmic properties, but instead in developing the skills that lead to quality.

[meheleventyone]

It’s also worth noting that the algorithm is not the only meta and posting consistently might be more or less expected depending on your audience, topic and so on.

[steveklabnik]

Absolutely, each audience has different expectations for what "quality" even means. Though I'm not really sure where "the algorithm" (I kinda hate how anthropomorphized it is) doesn't play into even in your scenario. Browse and Search are the two largest ways to grow traffic, and both are algorithmic to different degrees... unless you just mean like, growth is not a goal on YouTube for you, in which case, yeah you can just ignore all of this for sure.

But the language used around this topic is often implying that "the algorithm" will "punish you" if you like, post too often, post too little, post at different times... and I personally believe that that is empirically demonstrated to be not the case, and is also something YouTube (at least lately, I've only been paying close attention for a year or two) has never implied is important. And they've said things that imply that it's totally irrelevant, recommendation wise.

[411111111111111]

> But the language used around this topic is often implying that "the algorithm" will "punish you" if you like, post too often, post too little, post at different times...

These are all kinda true though?

- If you post too often your notifications get bundled, so now it's "x published two videos" instead of "x published that amazing video",

- if you post too little then people are less likely to recognize your username and thus just not click on it unless the creator really left an impression (this is probably the most dependent on the audience however).

- if you post at differing times you suddenly have to increase your audience significantly as your usual watchers won't be on YouTube at that moment, so you're effectively missing out on the initial upvotes by your fanbase until much later, reducing the chance of the video going into trending.

I mean they're not the only variables and they become pretty much non-issues as the channel grows, but all of these effects can be directly measured in clicks when the audience is still pretty small

[wingworks]

I would agree with that. I've been posting consistently every day, at the same time for 8 months and hasn't seen much return. But my vids are highly automated and pretty niche weather content, so wouldn't expect it to go crazy either. https://www.youtube.com/@aussiefromspace

I'm pretty happy, as a user that YouTube no longer pushes constancy over quality.

[pessimizer]

Another value in consistency is that when people are browsing aimlessly, they know when they can check your channel for new content. If I check some channel for new content a few times and there is none (and no indication of any coming), I will simply forget about it. Not intentionally. Someone else who posts predictably will have captured my idle time.

[steveklabnik]

As my sibling says, this may be how some users use youtube, generally these views ("channel page views" is the term used in the analytics) are a small minority of overall views.

In fact, this sort of behavior you're talking about is one of the reasons why "the algorithm" is a thing. If there's something you'd like to watch, but you've forgotten about it, it's YouTube's job to surface it to you, even if you've forgotten. This is one of the reasons why browse is the default view and not your subscribers feed, for example.

[jonny_eh]

This is not how most views occur, please don't hurt yourself to produce content regularly just to satisfy this edge case.