I’ve heard the argument that YouTube rewards consistency, but I am anything but consistent with my videos [1](every 1-3 months) and I haven’t noticed any negative consequences.
[1] YouTube.com/stuffmadehere
I’ve heard the argument that YouTube rewards consistency, but I am anything but consistent with my videos [1](every 1-3 months) and I haven’t noticed any negative consequences.
[1] YouTube.com/stuffmadehere
stuffmadehere is famous shit. Middle schoolers watch it in the classroom because the science teacher plays this. He has millions of subscribers & each show ( I refuse to call it video. Its a proper show. ) is very professionally done. It's so popular, if you were in India & you were speaking in Hindi, you would legit be on TV. Over there people are glued to content because some guy is explaining centrifugal force in hindi with some string & tincan. Meanwhile, look at your production values! Huge respect. Thanks for everything you do.
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.
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.
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
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.
I'm pretty happy, as a user that YouTube no longer pushes constancy over quality.
I know "The Algorithm" gets a lot of hate, but surfacing quality-over-quantity content is one place where recommendation algorithms excel over pure chronological feeds.
When I switched Twitter to chronological mode, I had to unfollow several people because they posted all day long. The most valuable Tweets I wanted to see were buried in the noise. The algorithmic feed is far from perfect, but it does a good job of highlighting posts I've missed from people I've interacted with previously. It takes some time to see the algorithm with enough likes and interactions, and it won't work if you never click the like button, but it's actually not half bad once it's up and running.
Reddit is another platform where quantity over quality prevails. If you click through to the post history of people who get content to the front page, it's basically a firehose of posts using recycled content and slightly altered headlines to as many subreddits as they can get away with. Eventually one of them clicks and rises to the top, but by that point the headline is often so mangled into clickbait that it doesn't accurately reflect the content of the article. Redditors don't really read articles, though, so it doesn't matter. Spam away.
I’ve seen this too. It seems like a lot of work and I’m unclear what the reward is. It isn’t like most of that content is pimping some brand of burger or something…
Do you find great content on youtube via YT's recommendations? I practically never have, everything I like I've stumbled upon on other sites where something is discussed and a channel is mentioned, or from someone who's work I like mention someone else. YT's recommendations have always been garbage for me, so I eventually just hid them completely because it's mostly noise.
I'm sure not-your-entire audience is interested in these things. Have you ever considered making a second channel where you go into more detail about the details of the technical problems / solutions you run into? The videos could be a lot less well produced, I'm just desperate for the info.
Thanks for all you do!
You have infinitely more Youtube experience than I do (love your videos, by the way!), but maybe those other categories are in your blindspot, just like they are not in my Youtube suggestions.
I'd expect vloggers, celebrities, podcasters, game streamers, political pundits, etc to benefit more from consistent uploads. More videos = more rolls of the viral dice, and furthermore frequency is required for building up fame, parasocial relationships, or just reliably filling the silence during commute.
Do normal users tend to just subscribe to a smaller number of channels, so they want channels that publish a lot?
In short, thank you for your awesome videos and take your time to keep the quality so high!
Keep up the good work!!
The mistake here is thinking that high quantity / low quality would lead to satisfied customers; not in the long term. When the quality is too low they'll unsub, simple as that.
My problem with YouTube videos in general is they don't cut to the point. They have too much fanfare around the point, stuff which does not interest me. I expect such behavior in advertising, not in content. To me, it lowers the quality of the content because it lowers the signal-to-noise ratio.
Now, there's enough content that the algorithm favors quality more, because YT can more easily keep a user by recommending similar content from a different creator of a higher quality.
.. again, no idea if that's true, but it does feel like it makes sense.
I think the consistency of popping out videos on the regular is about working the algorithm and grinding out content for views, and with those views: ad dollars. I've seen too many channels just turn into churning really awful, boring, long videos with little in the way of direction in the hopes you'll just mindlessly watch it. But I think the math just works out better (marginally) to pump out low-effort, low-value content more often than for example, what you are creating. And like I said: blame the algorithm.
I work with an org that funds and supports a really large community of young makers (mostly young people in their teens, but we also have folks in their mid-twenties and beyond). I'd love to talk to you about how you honed your skills. A lot of them watch your videos and I suspect they would be really interested in reading about how you honed your skills and achieved your currently glorious skillset. (I would be too!)
So I agree that quality is an important factor. While there has to be some weighting to favor the needs of advertisers in play, it's not as strong lately as it has been.
If anyone is giving you any kind of marketing or content distribution advice, and at no point do they mention product-market fit, or content quality, then they are selling you a dream.