YouTube plays a key role in the information operations and public relations of the Russian government, from the channels and networks of minor ministries to Russia Today, which claims the most views out of any news network on YouTube. Many of these videos are monetized, meaning YouTube places ads before or during the videos. We developed multiple models to estimate the Kremlin’s revenue from advertising on YouTube, as well as YouTube’s from placing ads before Kremlin content. We present all-time figures, based on a third party model, that place YouTube’s total at $60M and the Kremlin’s at $73M over a 12 year period. We highlight shortcomings with this model and build our own, though data limitations kept the scope to 2017–2018. Our first model estimates $27M to the Kremlin and $22M to YouTube over a 2 year period. Our second model accounts for variation in cost per view by country but makes more assumptions and estimates $7M to the Kremlin and $6M to YouTube over the previous two years.
YouTube generates revenue by charging advertisers to place their ads before videos uploaded to the tool. YouTube gives 55% of those charges to the owner of the video’s channel, while keeping the remainder. Channels must have at least 1,000 subscribers and 4,000 hours of watch time and complete an application to qualify. YouTube is ambiguous about whether a human reviews applications.
The scale of the Kremlin’s presence on YouTube is prolific. Kremlin properties on YouTube have, in total, garnered 30 billion views. Were this a single channel, it would rank as the 3rd most watched on YouTube of any kind.
The Kremlin’s media properties dominate broadcasting in Russia, both on YouTube and traditional TV. Four out of the top ten channels in Russia are wholly government owned with a fifth majority owned. These media entities are legally Unitary Enterprises, meaning all shares are owned by the government. The financial relationship between these enterprises and the rest of the government varies, though government officials play a direct role in editorial decisions. Peter Pomerantsev in Nothing Is True And Everything Is Possible delves into the means by which the Kremlin can influence various Russian broadcasters.
Google does not release data related to YouTube performance but an ecosystem of consultants providing benchmarks and estimates has developed to fill in the gaps. We relied on their models to estimate revenue.
We used a Wikipedia list to find broadcasters owned by the state. For each, we checked their website to verify their ownership. None were listed on Wikipedia incorrectly, but our list may omit other channels. We followed links to YouTube accounts on each channel’s website to ensure we were only considering official channels. We then included channels of individual shows on those channels. We included Channel 1 (not be confused with Russia 1), in which the government has a majority share. As the chairman of Channel 1’s board is the chief of Russia’s Foreign Intelligence Service, omitting Channel 1 would provide an incomplete view of Russian information operations. YouTube labels Channel 1 as state backed.
We then went to the official pages of the major ministries of the Russian Government and recorded any links to YouTube channels on those sites.
To test if a channel was monetized, we loaded three randomly sampled videos in an incognito window without a proxy and labeled any channels where an ad was served as monetized. Screenshots confirming the ads can be found here.
The YouTube Money Calculator on Influencer Marketing Hub offers a straightforward estimate of revenue to publishers for channels. The calculator accepts a channel id and returns data on subscribers, views, and a revenue estimate. The estimate is a function of subscribers, views, and videos, which we determined by finding that view count was not a perfect predictor of revenue.
The website resolves instantly. Parsing data on the date of videos and view count by video requires hundreds to thousands of calls to the YouTube API, which can take hours. This allows us to conclude that Influencer Marketing Hub does not calculate revenue on a per video basis.
To understand that complexity, we ingested data on all videos from monetized, state-owned Russian channels from the YouTube Data API v3. We assigned the view count for each video to the month in which the video was published. Although views do not necessarily all occur the same month as publication, month-to-month variation in cost per view was not significant enough to make a substantial difference.
Costs per view can vary among countries, though, and Russian media operations are decidedly international, with flagship channels in English, French, German, Spanish and Arabic. We wanted to account for these differences, but no data was available on cost per view by country. The closest proxy we could find was research by SolveMethod and Wordstream on cost per click for Google Searches. We assumed the ratio of two countries’ cost per click on Google Search was similar to the ratio of the same two countries’ cost per view on YouTube.
Since even this data was limited, we needed to choose a type country for each language so chose the one with the most YouTube users: Russia for Russian, The US for English (Russia owns several channels specifically targeting the UK but none are monetized), France for French, Mexico for Spanish, and Saudi Arabia for Arabic.
Influencer Marketing Hub Model (Page 1)
Global Cost Per View Model (2017–8 only)
Model Normalized By Language (2017–8 only)
YouTube and the Kremlin have both profited from their partnership to the tune of millions to tens of millions of dollars. The most common, but simplest, model puts the total at $73M for the Kremlin and $59M for YouTube over the span of 12 years. A model that accounted for changing costs per view over time found the Kremlin generating $27M and YouTube generating $22M from 2017–2018. Views of videos published in that period accounted for 44% of all views, suggesting a total close to that of the simple estimate. A third model aimed to account for differences in cost per view by country but required the assumption that the ratio of one country’s cost per click on Google Search to another’s was the same as the ratio of cost per view on YouTube. This was the most generous model, but still showed a relationship worth millions of dollars to both partners each year.