I have been gently experimenting with creating video clips from Kodsnack episodes. It felt like a natural step after we got regular episodes up on Youtube (and waited for the whole back catalog to get chewed through). Youtube seems to be at least a bit popular for podcasts, but discovery is of course a challenge as always. So the next question after a few weeks of extremely similar video stats was "how do people spread and discover podcasts on Youtube?"
Shorts were a natural guess, so I figured out a reasonable way to create them. First using Imovie, then - mainly in order to get transcription and captions - using Descript.
And once you have short videos in vertical format, it seems pretty natural and is pretty easy to also post those in other places. So, I started putting them on Instagram and Tiktok as well.
Now that I have been doing that for a few weeks, the early results are very interesting.
- Instagram and Youtube shorts are in the same ballpark for average clips, with Youtube usually coming out a few tens of views ahead. There are a few clips which have been much more popular than the others, and for those Youtube have significantly outpaced Instagram
- Tiktok, however, is the really fascinating bit. Regular clips - remember we are talking about the exact same video clips - outperform Youtube (and Instagram) by about a factor of ten. Not only that, the two latest clips I posted performed as average on Youtube and Instagram, but between three and four times better than average on Tiktok!
Either that last two ones was a complete fluke, or Tiktok keeps finding more people to show the clips. I was ready to write off the second most recent one as a fluke, but when the latest one did the same it gets harder to think so. My interest in how the next one will do just keeps growing with each video.
In any case, it seems clear which service is most worth prioritizing. I am very curious what the difference between the services are, because it does not seem completely intuitive that Tiktok would on average have a lot more Swedish viewers interested in code than the other two. I imagine it is the famous algorithm doing a better job than the other two, and perhaps users also simply spend a lot more time looking at clips and need more of them? And of course, a major part could definitely be the three counting "a view" with highly varying degrees of strictness.
Oh yeah, the whole goal of the exercise is supposedly to spread the word about the podcast and see if it might bring in another listener or two. That effect has decidedly not showed up yet, but I am willing to give it more time as long as creating the clips is as fun and quick as it is.