How popular is rutube.ru?
After the invasion of Ukraine Russia banned a bunch of Western social media, including Facebook, Instagram, Tiktok, and Twitter. A notable absence from this list is Youtube. Why so?
It’s not from any great love for the platform. Indeed, last August the government said that though they weren’t planning on banning it, Youtube nevertheless ‘deserves it’ for taking part in an ‘information war against [Russian] citizens’. There have been several calls to ban it, perhaps most notably from Evgeny Prigozhin who is bankrolling a currently ongoing competition to develop a Russian Youtube alternative.
Then why? One of the reasons, it seems, is that it would annoy people; Youtube is very popular. One is logistic: to accommodate the changes in traffic cutting off Youtube would lead to requires bolstering national internet infrastructure since its popularity means it requires a lot of computing power (amounting to 18% of mobile internet use last summer). And another — perhaps now outdated in light of the post-invasion crackdown — is that Russia got more than it lost in relying on Youtube, and in particular some say it was because Russia was able to use YT to disseminate English-language foreign policy via things like RT.
Each of these reasons makes evident something that is perhaps easy to overlook: these things matter. To have or not have Youtube in a country like Russia is not just a question of having or doing without cat videos. It’s a question of both the physical world (literally of rooms or buildings of hard disks and computers), and of domestic and foreign policy. Russia without Youtube is different from Russia with Youtube.
And zooming out even more it’s arguably an instructive episode in the ongoing fight for the internet. If a couple of decades ago all talk was of Russia turning West, and Putin was buddy buddies with Blair and Bush, now talk is of it turning East to Xi, and it’s natural to ask whether that will be reflected in its internet. Will it go for the China model — proprietary and geofenced apps that mesh very nicely with government — or with the Western model whereby innovation trickles out of Silicon Valley across the world until (data protection, labour) law gradually catches up with it or VC money runs out.
Russia arguably has one foot in each camp: it illegalized much big social media for hosting ‘extremist’ anti-government content; it has its own variants for some domains (notably VKontakte and search engine Yandex which aspires to be an everything app); but it still relies on Google for Youtube.
Given this importance, let’s ask: will Russia ban Youtube? I’ll present some reason to think: not easily.
Russia would ban Youtube, it should be clear, if it had an alternative: a local Youtube variant that people liked and could stand the traffic. So, we could answer the question affirmatively if we could find an alternative, whereas we get some evidence for a negative answer if we can’t find one.
A bit of googling reveals several options. Smotrim.ru is sort of like the BBC iPlayer, owned by the government, and containing a bunch of professional content that originally aired on TV. Yandex Zen hosts videos among other things, as does Facebook-alike VKontakte. And Rutube is often spoken of as the Russian Youtube.
I think we can exclude Smotrim and VKontakte. They just play different functional roles in the media ecosystem, just as the videos you watch on iPlayer and those you watch on Facebook or Twitter are different. Yandex Zen certainly seems similar, and googling you can see people compare them and wonder which is better.
But it’s rutube.ru which is the main and most visible alternative. Unlike Zen, it’s solely and dedicatedly a tube site (Zen is a general content aggregator), it looks like Youtube (ish), and it’s the one that people speak of when they speak of ‘Russian Youtube’. So, what is there to say about Rutube? Here are some things, starting with a curious story this day last week
You can’t see number of views
Last week the Rutube Telegram posted something curious:
Open view metrics appeared on RUTUBE
Russian video hosting has opened views for the premiere release of “Show Will”. The indicators are available to all users who access the site using a browser on a computer or smartphone. It is not yet possible to see their number through applications.
And so it is: surf to that show, and you’ll see the following:
56k watchers for the episode released 18 hours ago, most of which hours were night-time. Search around and you’ll see that last week’s episode got 1.72 million. So, a pretty popular show.
But, wait, why is that a story? It’s because for most videos you can’t see the view count. Pick an arbitrary other one:
No views! Apart from that, it’s the same, with all the same icons and details about when it was posted.
Note that it’s not just that view counts have been introduced. For the few weeks I’ve been looking at it, the relevant html has been like the below. Pardon my inexpert cropping, but below you see screenshots of the html first for Show of Will (by the way I don’t know if that translation is missing something — Will is the host or something’s name), and the second for the random video I picked. Look at ‘views_total’. For SoW, it’s 57k, so what is shown on screen. But for the random one, it’s 0. Since provably the video didn’t have zero views (I viewed it!) it must be that the 0s are hardcoded. And so on: all videos are like this, with hardcoded 0s.
It’s of course tempting to draw the obvious conclusion that the numbers aren’t being released because they are bad. But it would be nice if we could confirm that. In fact we can, because we can get a sense of how many views a video gets despite those hardcoded zeroes.
How to get a rough sense of view counts
It so happens that a few days before the Telegram story mentioned above was posted I found out something interesting about Rutube — you actually sort of can get view counts!
The reason for this is the following. A channel consists of playlists, and playlists, like individual videos, have a field for views in the source. While those counts don’t appear on the page as rendered in your browser, However, nevertheless this field is not set to zero — the actual counts are there. You just have to look at the html source.
Thus, for example, consider a channel of a person in the top five on the home page as of this writing. Click through, go to playlist (Плейлисты) consisting of four videos from February, and you’ll see the first playlist consisting of three videos in March.
Then, view source, and we see the following:
So there’s a field called ‘totalViews’ that has maybe kinda sorta plausible view counts for whole playlists. If you have the taste for it, you can strengthen the plausibility of this by seeing if the numbers look right: if smaller channels’ playlists have fewer views, longer playlists more, and so on. They do.
Moreover, we can do a bit more to confirm that that is indeed what the numbers mean, and leverage that information to get view counts for individual videos.
And we do so as following: find playlists that are updated regularly, for example with daily or weekly shows. Regularly record the totalViews number for the playlist, and compare the difference. When a new video is posted between two measurements, we can assume that the guts of the new views are for the new video.
Starting about ten days ago, that’s what I did. In particular — and somewhat arbitrarily, and in retrospect I could have picked better — I chose a daily news playlist by the broadcaster NTV; a popular entertainment program (МАСКА); and two current affairs series helmed by Margarita Simonyan (Ч. Т. Д) and Tigran Keosayan (Международной пилорамы). They broadcast respectively 4–5 times a day, once a week, theoretically once a week but it seems to have stopped, and once a week.
Here are the numbers, respectively for Simonyan, Keosayan, the daily ‘Today’ program, and The Mask. Bold indicates when a new program dropped: so there was a new Keosayan on the 15–16 time period, and also on the 22–23 time period; a new Mask on 16 and on 23.
And eyeballing it, a few things jump out: first, this is exactly what you’d think the numbers would be like were you tracking views. Secondly, the numbers aren’t great. This is especially so for Simonyan and Keosayan, the former of whom’s numbers are terrible.
If you know Russian media, you’ll know Simonyan is one of the most famous voices, often popping up in Twitter clips and so on. One might hope this reveals her in fact deep unpopularity among the people, but rather it reflects something else: she is, for whatever reason, splitting her audience.
We see this by looking at how she shares her videos. Here’s an example:
At the bottom, she has links to VKonkakte, Odnoklassiki (another social media site), Rutube, and Yandex Dzen, at each of which places she posts the video. And clicking through you see that on VKontakte, for example, the video, which does unbelievably terribly on Rutube, gets around 400k views:
It seems it’s the site and not the star that is, at least in part, the explanation.
(Before going on, note there’s perhaps more to look it: it seems that a big proportion of independent bloggers each have around 5k subscribers, which seems unususal, and if one googles ‘how much does rutube pay’ (in Russian) you’ll get at least one report that seems incongruous: a small channel getting much more than makes sense. But that’s for another time.)
What consequences can we draw from this exercise in data finding? I think the concept of network effects helps understand things here. Social media are notoriously subject to these network effects, the classic example of which is something like a Fax machine that becomes more valuable as more people use it. So with social media: the more users, the better experience for each user (through more choice, for example).
Often, in good subject to network effects, you need to get over local minima — have a burst of popularity that takes a resource from low use-low value to high use-high value. This is often hard to engineer, as witness failures like Google+, which, if anything in the world could have been bruteforced to succeed, surely would have been.
It seems to me that Rutube is stuck at a local minima, albeit it’s perhaps increasing. Apart from for the Mask, the numbers are not great, given this is an industry in which compensation is rewarded on a per thousand view basis. If even state-backed and famous (like Simonyan and Keosayan), or well-resourced channels (like NTV) are languishing in the low thousands (and inspecting other channels would confirm this), Rutube isn’t going to be attractive for bloggers and others seeking an exodus from Youtube. And since a streaming site like this fundamentally depends on getting that sort of user, Rutube isn’t going to be attractive full stop.
And so, then, a conclusion: if Russia will only ban Youtube if it has a viable alternative, and if the only viable alternative is Rutube, Russia won’t ban Youtube. These are both big ifs, but what we’ve seen goes at least some of the way to answering the question with which we began.
