On SystemCrafters, the three main topics covered are Emacs, Guix, and Scheme. I got curious about their relative popularity and figured some of you might find the results interesting. The data below includes all live videos with the exception of Crafter Hour videos and “part 2” videos. It was divided by whether or not the word “Emacs”, “Scheme”, or Guix was in the title. Some videos fell in multiple categories.
The first image below is a box-and-whisker plot, where the box represents the first to third quartile (25–75%), and the whiskers extend to 1.5x the interquartile range, ending at a data point. Any data points beyond that are shown individually. The red lines mark the median number of views. Scheme became a regular topic on the channel in 2023, and since video topics and view trends have been fairly stable since then, this plot only includes videos posted in 2023 or later.
The second plot is a scatter plot showing view counts over time, colored by topic. I capped views at 20,000, so if a video had 100,000 views it’s plotted as 20,000, because otherwise the scale would obscure most of the variation in the data.
Interesting, thanks for sharing! Not really surprising, I guess… the “market” for emacs is certainly going to be larger than that of guix, and it’s not really surprising that guile also is bigger than guix? I still think it’s valuable that these topics appear on the SC streams and videos, though, as I think life would be duller without them.
As a completely unrelated aside (but I feel worth mentioning in this context); I was listening to the podcast “The happiness lab” in the car today, and in the episode I was listening to, the guest laid out a good argument /against/ chasing popularity metrics (here; making more emacs videos because they are the most popular ones). In essence, he argued that it could turn an enjoyable experience into something that’s suddenly an energy drain, under the guise of what he calls “purpose mirages”. So what I’m saying, I guess, is that if the guix videos is what generates a feeling of purpose for David, more than the others, then just bring those guix vids on
Definitely agree with that last part. The fact that this channel delves into whatever David is passionate about rather than chasing views is a huge part of why it’s so good, and I hope he continues doing that. I’ve tried things I never would have otherwise.
Also, if you’re just chasing views, sadly the best way seem to be to just put out semi-contrarian hot-takes.
This also corresponds to sc-s-copy’s experience: Emacs livestreams usually hover around ~100-120 live viewers, whereas Guix/Scheme (Friday) livestreams are usually in the ~60-80s. I don’t have concrete numbers, but sc-s-copy sometimes posts [Meta] XXX viewers reached.
<leftpad> Talking about other personas. Ha, nerd!
Anyway…
Yes. It would also be a bit weird to have a Guix/Scheme course that pays the bills, but never actually have any Guix/Scheme content on the livestreams. That would be quite a disconnect.
Thanks for putting this together, it’s really interesting to look at! It’s a little sad that the biggest number of views on videos were almost 5 years ago, lol.
It’s extremely difficult for me to focus on making videos about anything that I don’t find interesting, so I doubt that will change
Yes, unfortunately the algorithm (and human psychology) tend to push YouTubers toward this kind of content. There’s definitely a benefit to growing a channel, but better to find an approach that is authentic and providing value even if it doesn’t rake in hundreds of thousands of views on the regular.
Well, that was about at the start of the Covid-19 pandemic, which probably brought a lot of people to the channel and into new habits. That time itself might be an outlier.
It’s not sad, at least not per my interpretation. The shrinking number of views is easily explained by the earlier videos casting a much wider net. They were accessible to folks like me, who knew Emacs only as that ugly blinding white relic from the nineties with a tangle of incomprehensible keybindings, hadn’t even heard of Guix, and thought Lisp was some arcane mess of parentheses. In the past five years, all of those things have changed for me, especially my view of Emacs, thanks to those earlier videos.
The newer content is a bit more niche. Without the foundation from those earlier videos, things like Howm, Denote, and Consult, topics covered in the most recent Friday stream, would be meaningless to me. And I think that’s been a general trend: the more recent videos are going to attract people who have either been committed to learning this stuff from the channel’s earlier history, or are interested in diving into some of these topics because they already use the surrounding tools. This naturally means a smaller, but more dedicated, audience. That’s not a bad thing though. Fewer views don’t necessarily mean a smaller impact. Growth in viewer knowledge is a huge impact, and the deeper discussions also help strengthen the core of the community. All that said, my guess is that if you went back to newbie-focused videos, those videos would probably rack up higher views, but that’s a different kind of growth.
One other thing I’ll mention is that the earlier videos seem to have been advertised a bit more. I found this channel through a Reddit link I believe. I would wager that the majority of folks who come to the livestreams right now are already followers of your channel. Perhaps a bit more outreach could help bring in people who’d enjoy this content and community but haven’t stumbled on it yet.
